WO2024109247A1 - 图像处理方法及装置 - Google Patents

图像处理方法及装置 Download PDF

Info

Publication number
WO2024109247A1
WO2024109247A1 PCT/CN2023/116594 CN2023116594W WO2024109247A1 WO 2024109247 A1 WO2024109247 A1 WO 2024109247A1 CN 2023116594 W CN2023116594 W CN 2023116594W WO 2024109247 A1 WO2024109247 A1 WO 2024109247A1
Authority
WO
WIPO (PCT)
Prior art keywords
shadow
image
moving object
pixel
main image
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Ceased
Application number
PCT/CN2023/116594
Other languages
English (en)
French (fr)
Inventor
江春平
刘智超
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Honor Device Co Ltd
Original Assignee
Honor Device Co Ltd
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Honor Device Co Ltd filed Critical Honor Device Co Ltd
Priority to CN202380062845.0A priority Critical patent/CN119790435A/zh
Priority to EP23893340.2A priority patent/EP4498326A4/en
Priority to US18/860,941 priority patent/US20250356552A1/en
Publication of WO2024109247A1 publication Critical patent/WO2024109247A1/zh
Anticipated expiration legal-status Critical
Ceased legal-status Critical Current

Links

Classifications

    • G—PHYSICS
    • G06—COMPUTING OR CALCULATING; COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T15/00—Three-dimensional [3D] image rendering
    • G06T15/50—Lighting effects
    • G06T15/60—Shadow generation
    • G—PHYSICS
    • G06—COMPUTING OR CALCULATING; COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T7/00—Image analysis
    • G06T7/20—Analysis of motion
    • G06T7/246—Analysis of motion using feature-based methods, e.g. the tracking of corners or segments
    • G—PHYSICS
    • G06—COMPUTING OR CALCULATING; COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T11/00—Two-dimensional [2D] image generation
    • G06T11/60—Creating or editing images; Combining images with text
    • G—PHYSICS
    • G06—COMPUTING OR CALCULATING; COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T15/00—Three-dimensional [3D] image rendering
    • G06T15/50—Lighting effects
    • G—PHYSICS
    • G06—COMPUTING OR CALCULATING; COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T3/00—Geometric image transformations in the plane of the image
    • G06T3/18—Image warping, e.g. rearranging pixels individually
    • G—PHYSICS
    • G06—COMPUTING OR CALCULATING; COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T7/00—Image analysis
    • G06T7/20—Analysis of motion
    • G06T7/246—Analysis of motion using feature-based methods, e.g. the tracking of corners or segments
    • G06T7/248—Analysis of motion using feature-based methods, e.g. the tracking of corners or segments involving reference images or patches
    • G—PHYSICS
    • G06—COMPUTING OR CALCULATING; COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T2207/00—Indexing scheme for image analysis or image enhancement
    • G06T2207/10—Image acquisition modality
    • G06T2207/10016—Video; Image sequence
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04N—PICTORIAL COMMUNICATION, e.g. TELEVISION
    • H04N7/00—Television systems
    • H04N7/01—Conversion of standards, e.g. involving analogue television standards or digital television standards processed at pixel level
    • H04N7/0135—Conversion of standards, e.g. involving analogue television standards or digital television standards processed at pixel level involving interpolation processes
    • H04N7/0137—Conversion of standards, e.g. involving analogue television standards or digital television standards processed at pixel level involving interpolation processes dependent on presence/absence of motion, e.g. of motion zones

Definitions

  • the present application relates to the field of image processing technology, and in particular to an image processing method and device.
  • Predictive frame technology is a mainstream technology in image rendering.
  • Electronic devices can use predictive frame technology to reduce the cost of image rendering.
  • electronic devices add shadow rendering.
  • Predictive frame technology causes shadow jitter problems in predicted frames, increasing the difference between predicted frames and real images, greatly reducing user experience.
  • Real images are images generated by electronic devices according to rendering logic, and real images are used to generate predicted frames.
  • the present application provides an image processing method and device, the purpose of which is to solve the problem of shadow jitter in the predicted frame, reduce the difference between the predicted frame and the real image, and improve the user experience.
  • the present application provides the following technical solutions:
  • the present application provides an image processing method, which is applied to an electronic device, and the method includes: obtaining a first main image from a first real image and obtaining a second main image from a second real image, wherein the first real image and the second real image are images rendered by the electronic device; calculating motion vectors for the first main image and the second main image to obtain motion vectors; distorting the environment in the target main image and the moving object in the target main image according to the motion vectors to generate the environment in the predicted frame and the moving object in the predicted frame, wherein the target main image is a frame of the first main image and the second main image; and reusing the image data of the shadow of the moving object in the target main image in the predicted frame to generate the shadow of the moving object in the predicted frame.
  • the shadow of the moving object in the predicted frame can be extracted from the environment in the predicted frame, so that the shadow of the moving object in the predicted frame can be generated independently, and the shadow of the moving object in the predicted frame no longer depends on the motion vector, but is generated by reusing the image data of the shadow of the moving object in the target main image in the predicted frame, thereby reducing the shadow jitter problem existing in the predicted frame, reducing the difference between the predicted frame and the real image, and improving the user experience.
  • the environment in the target main image and the moving object in the target main image are warped according to the motion vector to generate the environment in the predicted frame and the moving object in the predicted frame, and the image data of the shadow of the moving object in the target main image is reused in the predicted frame to generate the shadow of the moving object in the predicted frame, including: according to the motion vector, the environment in the target main image and the moving object are warped to generate a first intermediate image, and the first intermediate image includes the environment and the moving object; in the first intermediate image, the image data of the shadow of the moving object is reused so that the first intermediate image includes the environment, the moving object, and the shadow of the moving object.
  • the first intermediate image After obtaining the first intermediate image including the environment, the moving object, and the shadow of the moving object, the first intermediate image is completed to obtain a second intermediate image; then the second intermediate image is post-processed to fuse the user interface image corresponding to the target main image and the processed second intermediate image to obtain a predicted frame. Because the moving object and the shadow of the moving object have not changed after obtaining the first intermediate image, the moving object in the first intermediate image can be the moving object in the predicted frame, and the first intermediate image The shadow of the moving object in the image can be the shadow of the moving object in the prediction frame.
  • the environment in the first intermediate image is completed during the image completion process, so the prediction frame includes the completed environment in the first intermediate image, that is, the environment in the second intermediate image.
  • reusing the image data of the shadow of the moving object in the target main image in the predicted frame includes: for each pixel in the predicted frame, according to the position of the shadow of the moving object in the target main image, determining whether the pixel is located in the shadow of the moving object in the target main image; in determining that the pixel is located in the shadow of the moving object in the target main image, reusing the pixel value of the pixel in the target main image, the pixel value of the pixel in the target main image is the image data of the pixel in the target main image; in determining that the pixel is not located in the shadow of the moving object in the target main image, keeping the pixel value of the pixel unchanged.
  • the pixel value of the pixel in the target main image is reused, so that both the position of the shadow of the moving object in the target main image and the image data of the shadow of the moving object in the target main image are reused, so that the shadow of the moving object in the target main image is merged into the predicted frame, avoiding the continuous use of the motion vector used by the environment to generate the shadow of the moving object in the predicted frame, reducing the shadow jitter problem existing in the predicted frame, reducing the difference between the predicted frame and the real image, and improving the user experience.
  • the prediction frame reuses the position of the shadow of the moving object in the target main image
  • the position of the shadow of the moving object in the prediction frame is the same as its position in the target main image.
  • the shadow of the moving object has the same shape, color, and brightness in the prediction frame and the target main image.
  • the motion vector is used to record the moving speed and moving direction of each pixel in the target main image, and according to the motion vector, the environment in the target main image and the moving object in the target main image are distorted, and the environment in the predicted frame and the moving object in the predicted frame are generated, including: according to the moving speed and moving direction of each pixel in the target main image, each pixel in the target main image is moved; in the process of moving the pixel, according to the position of the shadow of the moving object in the target main image, whether the pixel is located in the shadow of the moving object in the predicted frame is determined; in the case where it is determined that the pixel is located in the shadow of the moving object in the predicted frame, the pixel value of the pixel is set to a preset value; in the case where it is determined that the pixel is not located in the shadow of the moving object in the predicted frame, the pixel value of the pixel is set to the pixel value of the pixel in the target main image.
  • the shadow of the moving object in the predicted frame is distinguished from other objects in the predicted frame by the preset value, so that the shadow of the moving object in the predicted frame is removed by the preset value and the position of the shadow of the moving object can be marked, so as to reserve the shadow of the moving object for the predicted frame, so that the predicted frame can reuse the image data of the shadow of the moving object in the target main image.
  • the preset value includes a preset transparency
  • the preset transparency is different from the transparency of other pixels in the target main image, and is distinguished by the preset transparency.
  • the process of determining the position of the shadow of a moving object in a target main image includes: comparing the pixels to be compared that are in the same position in the environment shadow texture image and the moving object shadow texture image, the environment shadow texture image is generated based on the image data of the environment shadow in the target main image, and the moving object shadow texture image is generated based on the image data of the shadow of the moving object in the target main image; when it is determined that the pixel to be compared is in the shadow area of the moving object but not in the environment shadow, marking the pixel to be compared as being in the shadow area of the moving object; when it is determined that the pixel to be compared is in the shadow area of the moving object and in the environment shadow, marking the pixel to be compared as not being in the shadow area of the moving object; determining the position of the shadow of the moving object in the target main image based on the pixel to be compared that is marked as being in the shadow area of the moving object, the image data of the shadow of the moving object in the target main image
  • the image data of the shadow of the moving object in the target main image is the same as the image data of the shadow of the moving object in the target main image.
  • the position of the shadow of the moving object in the target main image can be determined according to the principle of "when the shadow of the moving object is in the environment shadow, the shadow of the moving object is not presented in the target main image", in preparation for determining the position of the shadow of the moving object in the predicted frame, so that the image data of the shadow of the moving object can be reused.
  • the pixel value of the pixel to be compared in the moving object shadow texture image is greater than the pixel value of the pixel to be compared in the environment shadow texture image, it is determined that the pixel to be compared is in the shadow area of the moving object but not in the environment shadow; if the pixel value of the pixel to be compared in the moving object shadow texture image is less than or equal to the pixel value of the pixel to be compared in the environment shadow texture image, it is determined that the pixel to be compared is in the shadow area of the moving object and in the environment shadow.
  • the pixel value of the pixel to be compared includes at least one of the R channel value, the G channel value, and the B channel value of the pixel to be compared.
  • the pixel value of the pixel to be compared includes one of the R channel value, the G channel value, and the B channel value, which can save storage space; in some examples, the pixel value of the pixel to be compared includes two or three of the R channel value, the G channel value, and the B channel value, then when comparing the pixel values, the pixel values of the pixel to be compared can be first fused (such as weighted average) to obtain a numerical value, and then the numerical values are compared, thereby comprehensively considering the influence of different channel values.
  • the method before obtaining the first main image from the first real image and obtaining the second main image from the second real image, the method further includes: when rendering the real image corresponding to the target main image, obtaining image data of the environmental shadow and image data of the shadow of the moving object, the image data of the environmental shadow and the image data of the shadow of the moving object are used to determine the position of the shadow of the moving object in the target main image, the position of the shadow of the moving object in the target main image and the image data of the shadow of the moving object are used to obtain the image data of the shadow of the moving object to be reused in the target main image.
  • the image data of the environmental shadow and the image data of the shadow of the moving object are intercepted to prepare for reuse.
  • obtaining the image data of the environment shadow and the image data of the shadow of the moving object includes: rendering the environment shadow of the real image to generate the image data of the environment shadow, and the image data of the environment shadow is stored in the first frame buffer; creating the second frame buffer and the third frame buffer; copying the image data of the environment shadow stored in the first frame buffer to the second frame buffer; changing the frame buffer bound to the real image from the first frame buffer to the third frame buffer; rendering the shadow of the moving object of the real image to generate the image data of the shadow of the moving object, and the image data of the shadow of the moving object is stored in the third frame buffer.
  • the image data of the environment shadow and the image data of the shadow of the moving object are stored in different frame buffers respectively, so as to prevent the two parts of the image data from overlapping and causing the problem of being difficult to distinguish, thereby facilitating the use of the image data of the environment shadow and the image data of the shadow of the moving object.
  • obtaining the image data of the environment shadow and the image data of the shadow of the moving object also includes: after rendering the shadow of the moving object of the real image, changing the frame buffer bound to the real image from the third frame buffer to the first frame buffer; rendering the shadow of the moving object of the real image again to generate the image data of the shadow of the moving object, and storing the image data of the shadow of the moving object in the first frame buffer.
  • the image data of the environment shadow and the image data of the shadow of the moving object can be stored in the first frame buffer bound to the real image, so as to ensure the accuracy of the image data stored in the first frame buffer and prevent the real image (mainly the target main image) from lacking a shadow.
  • copying the image data of the ambient shadow stored in the first frame buffer to the second frame buffer includes: calling the glBlitFramebuffer function to copy the image data of the ambient shadow stored in the first frame buffer to the second frame buffer.
  • changing the frame buffer to which the real image is bound from the first frame buffer to the third frame buffer includes: calling the glBindFrameBuffer function to change the frame buffer to which the real image is bound from the first frame buffer to the third frame buffer.
  • the method further includes: after generating the shadow of the moving object in the predicted frame, performing image completion on the predicted frame.
  • the present application provides an electronic device comprising: one or more processors and a memory; the memory is used to store computer program code, the computer program code includes computer instructions, and when the one or more processors execute the computer instructions, the electronic device executes the image processing method described in the first aspect.
  • the present application provides a computer-readable storage medium for storing a computer program.
  • the computer program When executed, it is specifically used to implement the image processing method as described in the first aspect.
  • FIG1 is a schematic diagram of generating a prediction frame provided by the present application.
  • FIG2 is a hardware structure diagram of an electronic device provided by the present application.
  • FIG3 is a flowchart of generating a real image provided by the present application.
  • FIG4 is a flowchart of generating a prediction frame provided by the present application.
  • FIG. 5 is a flowchart of obtaining image data of environmental shadows and image data of shadows of moving objects in a real image provided by the present application.
  • FIG6 is a flow chart of determining the position of the shadow of a moving object in a real image provided by the present application
  • FIG7 is a flowchart of removing the shadow of a moving object generated according to a motion vector provided by the present application.
  • FIG8 is a flow chart of multiplexing image data of shadows of moving objects in real images provided by the present application.
  • FIG9 is a schematic diagram of multiplexed shadows provided in the present application.
  • one or more refers to one, two or more; “and/or” describes the association relationship of the associated objects, indicating that three relationships may exist; for example, A and/or B can represent: A exists alone, A and B exist at the same time, and B exists alone, where A and B can be singular or plural.
  • the character “/” generally indicates that the objects associated before and after are in an "or” relationship.
  • references to "one embodiment” or “some embodiments” in this specification mean that one or more embodiments of the present application include a particular feature, structure or characteristic described in conjunction with the embodiment.
  • the phrases “in one embodiment”, “in some embodiments”, “in some other embodiments”, “in some other embodiments”, etc. appearing in different places in this specification do not necessarily refer to the same embodiment, but mean “one or more but not all embodiments”.
  • the terms “include”, “comprising”, “having” and their variations all mean “including but not limited to”, unless otherwise specifically emphasized.
  • the electronic device can display multimedia stream files to the user through a display screen, so as to provide the user with a rich visual experience.
  • the multimedia stream may include multiple frames of images.
  • the video stream may include N frames of images, such as the 1st frame, the 2nd frame, and ... the Nth frame.
  • the electronic device can display the 1st frame, the 2nd frame, ... the Nth frame on the display screen respectively.
  • An application that wants to play the video stream can issue rendering commands for different frame images to the electronic device.
  • the electronic device can render each frame image according to these rendering commands, obtain the rendering results (such as image data) of each frame image, and display based on the rendering results.
  • Figure 1 is a schematic diagram of generating a prediction frame.
  • the process of generating a prediction frame is: the electronic device performs user interface (UI) separation on a first real image and a second real image to obtain a first main image of the first real image and a first UI image of the first real image, a second main image of the second real image and a second UI image of the second real image, wherein the first real image is a previous frame image of the second real image, the UI image includes a UI interface in the real image, and the main image includes objects such as the environment, environmental shadows, moving objects, and shadows of moving objects in the real image.
  • UI user interface
  • the electronic device calculates a motion vector (MV) based on the first main image and the second main image to obtain the motion vector, where the motion vector indicates the moving direction and speed of each object in the second main image; performs image warping (ImageWarp) on the second main image based on the motion vector to obtain a first intermediate image, wherein ImageWarp uses the motion vector to move each object in the second main image; since the moving speeds and directions of each object in the second main image are different, image data is missing in the first intermediate image, so after obtaining the first intermediate image, image completion (HoleFilling) is performed on the first intermediate image to obtain the second intermediate image; the second intermediate image is fused with the second UI image (i.e., the second UI image is attached to the second intermediate image) to generate a prediction frame, thereby generating a prediction frame based on the first real image and the second real image.
  • MV motion vector
  • ImageWarp image warping
  • the predicted frame may be an intermediate frame between the first real image and the second real image; in some examples, the predicted frame may be the next frame of the second real image.
  • the electronic device may perform ImageWarp on the first main image according to the motion vector to obtain the first intermediate image.
  • the second intermediate image and the first UI image are fused to generate the predicted frame.
  • the electronic device can add shadow rendering to the image rendering, that is, for the moving objects in the first real image and the second real image, the electronic device renders a shadow for the moving object, and the shadow moves with the moving object.
  • the shadow of the character is rendered, and when the character moves, the shadow of the character moves with the character.
  • the predicted frame of the character is generated, the predicted frame also has a shadow.
  • the image also includes the environment.
  • the environment refers to the scene where the moving object is located.
  • the electronic device also renders a shadow on the environment, which is called the environmental shadow.
  • the shadow of the moving object moves according to the moving direction and speed of the environment.
  • the present application provides an image processing method, in which, in the process of generating a prediction frame, the image data of the shadow of a moving object in a real image is reused into the prediction frame to generate the shadow of the moving object in the prediction frame, and the moving object, environment and environment shadow in the prediction frame are generated by motion vectors, a main image of a frame of real image and a UI image of the real image.
  • the image data of the shadow of the moving object in the first real image or the second real image is reused.
  • the shadow of the moving object generated according to the motion vector is removed during the process of generating the predicted frame.
  • the electronic device renders the real image
  • the image data of the shadow of the moving object in the real image and the position of the shadow of the moving object in the real image are obtained, wherein the position of the shadow of the moving object in the real image is to determine the position of the shadow of the moving object in the predicted frame, and the image data of the shadow of the moving object in the real image is for reuse in the predicted frame.
  • the shadow of the moving object in the predicted frame no longer depends on the moving direction and moving speed of the environment, but is generated by reusing the image data of the shadow of the moving object in the real image in the predicted frame, thereby reducing the shadow jitter problem existing in the predicted frame, reducing the difference between the predicted frame and the real image, and improving the user experience.
  • the main image of the real image includes the shadow of the moving object, and reusing the image data of the shadow of the moving object in the real image may refer to reusing the image data of the shadow of the moving object in the main image of the real image, and the position of the shadow of the moving object in the real image may refer to the position of the shadow of the moving object in the main image.
  • the image data of the shadow of the moving object in the reused main image may also be referred to as the image data of the shadow of the moving object in the main image.
  • the shadow rendering logic of the application is: when the shadow of the moving object is located in the environmental shadow, the shadow of the moving object is not presented in the real image. Therefore, when determining the position of the shadow of the moving object in the real image, there is a situation where the shadow that is located in both the moving object shadow and the environmental shadow is removed. Therefore, in some examples, the image data of the shadow of the moving object in the main image is the image data that is located in the shadow of the moving object but not in the environmental shadow, that is, partial image data of the shadow of the moving object in the main image; for example, in the following step S306, all image data of the shadow of the moving object is obtained, and the image data of the shadow of the moving object in the main image is partial image data of the total image data obtained in step S306. In some examples, the image data of the shadow of the moving object in the main image is the same as the total image data obtained in step S306.
  • the electronic device can be a device that can provide network access.
  • the electronic device can be a mobile phone, a tablet computer, a desktop, a laptop, a notebook computer, an ultra-mobile personal computer (Ultra-mobile Personal Computer, UMPC), a handheld computer, a netbook, a personal digital assistant (Personal Digital Assistant, PDA), a wearable electronic device, a smart watch and other devices.
  • the present application does not specifically limit the specific form of the electronic device.
  • the electronic device can have a display function. For example, the electronic device can generate a predicted frame based on the first real image and the second real image, display the predicted frame, and display the first real image and the second real image.
  • the electronic device may include: a processor, an external memory interface, an internal memory, a Universal Serial Bus (USB) interface, a charging management module, a power management module, a battery, an antenna 1, Antenna 2, mobile communication module, wireless communication module, sensor module, button, motor, indicator, camera, display screen, and Subscriber Identity Module (SIM) card interface, etc.
  • the audio module may include a speaker, a receiver, a microphone, an earphone interface, etc.
  • the sensor module may include a pressure sensor, a gyroscope sensor, an air pressure sensor, a magnetic sensor, an acceleration sensor, a distance sensor, a proximity light sensor, a fingerprint sensor, a temperature sensor, a touch sensor, an ambient light sensor, a bone conduction sensor, etc.
  • the structures illustrated in the embodiments of the present application do not constitute specific limitations on the electronic device.
  • the electronic device may include more or fewer components than shown in the figure, or combine certain components, or split certain components, or arrange the components differently.
  • the illustrated components may be implemented in hardware, software, or a combination of software and hardware.
  • the processor may include one or more processing units, for example, the processor may include an application processor (AP), a modem processor, a graphics processor (GPU), an image signal processor (ISP), a controller, a video codec, a digital signal processor (DSP), a baseband processor, and/or a neural network processor (NPU). Different processing units may be independent devices or integrated into one or more processors.
  • the processor is the nerve center and command center of the electronic device.
  • the controller may generate an operation control signal according to the instruction opcode and the timing signal to complete the control of fetching and executing instructions.
  • the external memory interface can be used to connect an external memory card, such as a Micro SD card, to expand the storage capacity of the electronic device.
  • the external memory card communicates with the processor through the external memory interface to implement the data storage function.
  • the internal memory can be used to store computer executable program code, which includes instructions.
  • the processor executes various functional applications and data processing of the electronic device by running the instructions stored in the internal memory. For example, in an embodiment of the present application, the processor enables the electronic device to execute the image processing method provided in the embodiment of the present application by running the instructions stored in the internal memory.
  • GPU is a microprocessor for image processing, connecting the display screen and application processor. GPU is used to perform mathematical and geometric calculations for image rendering.
  • the processor may include one or more GPUs, which execute program instructions to generate or change display information.
  • the display screen is used to display images, videos, etc.
  • the display screen includes a display panel.
  • the display panel can be a liquid crystal display (LCD), an organic light-emitting diode (OLED), an active-matrix organic light-emitting diode or an active-matrix organic light-emitting diode (AMOLED), a flexible light-emitting diode (FLED), Miniled, MicroLed, Micro-oled, quantum dot light-emitting diodes (QLED), etc.
  • the electronic device may include 1 or N display screens, where N is a positive integer greater than 1.
  • GUIs graphical user interfaces
  • a control is a GUI element, which is a software component included in an application program, and controls all data processed by the application program and interactive operations on these data. Users can interact with the control through direct manipulation to read or edit relevant information of the application program.
  • controls can include visual interface elements such as icons, buttons, menus, tabs, text boxes, dialog boxes, status bars, navigation bars, widgets, etc.
  • the display screen can display a video file containing a specific action, etc.
  • the processor includes a CPU and a GPU.
  • the CPU can generate a rendering instruction stream according to the rendering logic of the application program, and send the rendering instruction stream to the GPU so that the GPU can perform corresponding rendering according to the rendering instruction stream.
  • the CPU can send rendering instructions of the rendering instruction stream to the GPU one by one.
  • the CPU can send rendering instructions to the GPU again after receiving a response from the GPU; in some examples, the CPU does not need to wait for the response of the GPU, and continues to send the next rendering instruction after completing the sending of one rendering instruction, depending on the CPU's setting of the rendering instruction.
  • the GPU when the GPU is performing the rendering of the real image, the GPU can obtain the image data of the shadow of the moving object in the real image, so as to reuse this part of the image data when generating the predicted frame.
  • FIG3 it shows the process of generating the real image in the image processing method provided in the embodiment of the present application.
  • the image data of the shadow of the moving object in the real image is obtained, which can include the following steps:
  • the rendering instruction stream includes at least one rendering instruction, and the rendering instruction instructs the GPU to perform corresponding rendering, such as instructing the GPU to render a moving object in a real image, instructing the GPU to render an environment in a real image, and so on.
  • the application when the application turns on shadow rendering, the application calls at least one of a specific program, a specific resource, and a specific rendering instruction, so that the judgment condition can record the information called when the application turns on shadow rendering.
  • the judgment condition can record the label of the specific program to identify the specific program through the label, for example, the label of the specific program can be the program name of the specific program;
  • the judgment condition can record the resource label of the specific resource to identify the specific resource through the resource label, for example, the resource label can be the resource name, etc.; if the application calls a specific instruction when turning on shadow rendering, the judgment condition can record the specific instruction.
  • step S107 perform corresponding rendering according to the rendering instruction stream to obtain all image data of the real image, and complete the rendering of the environmental shadow and the shadow of the moving object through steps S103 to S106.
  • a frame of real image may also include moving objects, environment, and UI interface, etc. Therefore, in some examples, the rendering here is mainly the rendering of the image except the environmental shadow and the shadow of the moving object, such as the rendering of the moving object, environment, UI interface, etc. In some examples, the rendering of the moving object and the environment may be before the shadow rendering, so the rendering of step S107 may be the rendering of the image except the environmental shadow, the shadow of the moving object, the moving object and the environment, such as the rendering of the UI interface.
  • step S105 may be executed to render the shadow of the moving object.
  • step S107 may be executed to perform corresponding rendering according to the rendering instruction stream to complete the rendering of the real image.
  • S108 Output the real image according to all the image data of the real image. After obtaining all the image data of the real image, generate the real image according to all the image data of the real image, and display the real image.
  • FIG3 shows a process of generating a prediction frame, in which the image data of the environment shadow and image data of the shadow of the moving object are used, and the following steps may be included:
  • S202 MV calculation is performed on the first main image of the first real image and the second main image of the second real image to obtain a motion vector. That is, MV calculation is performed on the first main image and the second main image to obtain a motion vector, which indicates the moving direction and moving speed of each object in the second main image.
  • step S203 determine whether to enable shadow rendering, if yes, execute step S204, if no, execute step S208.
  • the determination of whether to enable shadow rendering can refer to the above step S102, which will not be repeated here.
  • S204 Determine a region where the shadow of the moving object in the real image is located according to the image data of the environmental shadow and the image data of the shadow of the moving object.
  • the image distortion of the second main image according to the motion vector mainly uses the motion vector to move each object in the second main image, specifically, for any object in the second main image, the object is moved according to the moving direction and moving speed of the object in the motion vector.
  • the second main image includes the environment, the shadow of the environment, the moving object and the shadow of the moving object.
  • the moving object and the environment are first separated from the second main image. This separation separates the environment and the shadow of the moving object from the moving object, that is, the second main image is separated into two parts, the first part includes the environment, the shadow of the environment and the shadow of the moving object, and the second part includes the moving object.
  • the moving direction and moving speed of each object in the first part of the motion vector are used to move each object in the first part; and the moving direction and moving speed of the moving object in the second part of the motion vector are used to move the moving object.
  • it is prohibited to reuse the image data of the shadow of the moving object in the real image.
  • the shadow of the moving object in the first intermediate image is removed, and the shadow of the moving object can be removed in the process of generating the first intermediate image.
  • the shadow of the moving object in the first intermediate image is removed.
  • the first intermediate image obtained in step S205 is an image with an environment, environmental shadows, and a moving object, but the shadow of the moving object is removed.
  • the purpose of removing the shadow of the moving object is to mark the position of the shadow of the moving object in the first intermediate image.
  • the image data of the shadow of the moving object in the real image is not used in the process of generating the first intermediate image, so that a to-be-completed area with pixel values to be completed is reserved at the marked position.
  • the position of the shadow of the moving object marked in the first intermediate image is obtained by moving the shadow of the moving object in the real image according to the motion vector, so the position of the shadow of the moving object marked in the first intermediate image is different from its position in the real image.
  • multiplexing the image data of the shadow of the moving object in the second main image not only multiplexes the image data but also multiplexes the position of the shadow of the moving object in the second main image, so that the shadow of the moving object in the second main image is pasted in the first intermediate image.
  • the position of the shadow of the moving object marked in the first intermediate image is different from its position in the second main image. Therefore, after multiplexing the image data of the shadow of the moving object, the position of the shadow of the moving object marked in the first intermediate image still has a to-be-completed area whose pixel values need to be completed. The pixel values of each pixel in the to-be-completed area are obtained through image completion processing.
  • the moving speed and moving direction of each object in the second main image may be different, resulting in missing image data in the first intermediate image.
  • the area with missing image data can be called the area to be completed.
  • the pixel value of each pixel in the area to be completed is obtained through image completion processing.
  • a deep learning network may be used to complete the pixel values of each pixel in the area to be completed.
  • a traditional image restoration operation may be used to complete the pixel values of each pixel in the completed area.
  • a mask image of the first intermediate image is used to complete the pixel values of each pixel in the area to be completed, so as to determine some static pixels corresponding to the pixels to be completed, and the pixel values of these static pixels are used to determine the pixel values of the pixels to be completed.
  • the pixels to be filled there are three types of pixels, one is the pixels to be filled, one is the static pixels, and one is the dynamic pixels.
  • the pixels to be filled refer to the pixels that need to be filled in the area to be filled in the mask image (the same as the first intermediate image).
  • the moving object and the shadow of the moving object correspond to a region to be filled respectively.
  • the region to be filled can be a blank area.
  • the pixels in the blank area near the moving object are the pixels to be filled, and the pixels in the blank area near the shadow of the moving object are the pixels to be filled;
  • the static pixels refer to the pixels corresponding to the environment in the mask image that is in a static state, and their pixel values can be used to determine the pixel values of the pixels to be filled;
  • the dynamic pixels refer to the pixels corresponding to the objects in motion (or called moving objects) in the mask image, and their pixel values cannot be used to determine the pixel values of the pixels to be filled.
  • the static pixel point is a valid reference point for completing the pixel point to be complemented
  • the dynamic pixel point is an invalid reference point for completing the pixel point to be complemented.
  • the grayscale value of the pixel point to be complemented is a first grayscale value
  • the grayscale value of the static pixel point is a second grayscale value
  • the grayscale value of the dynamic pixel point is a third grayscale value.
  • the first grayscale value, the second grayscale value and the third grayscale value are not equal to each other.
  • the grayscale value of the static pixel point is 0, the grayscale value of the dynamic pixel point is 1, and the grayscale value of the pixel point to be complemented is 0.5.
  • the electronic device determines a group of valid static pixels corresponding to the pixel to be filled in the mask image, and determines the pixel value of the pixel to be filled based on the pixel values of the group of valid static pixels in the first intermediate image.
  • One possible implementation is to convert the pixel values of the group of valid static pixels in the first intermediate image into pixel values. The weighted average of the values is used as the pixel value of the pixel to be filled.
  • the electronic device fills the first intermediate image with pixel values according to the pixel values of each pixel to be filled to obtain the second intermediate image.
  • a group of valid static pixels corresponding to a pixel to be complemented has the same pixel spacing as the pixel to be complemented, and compared with other static pixels (which can be called invalid static pixels), the pixel spacing between this group of valid static pixels and the pixel to be complemented is the smallest.
  • the post-processing mainly involves performing image filtering on the second intermediate image.
  • the shadow of the moving object in the predicted frame can be extracted from the environment of the predicted frame, so that the shadow of the moving object in the predicted frame no longer depends on the motion vector generation, but is generated by reusing the image data of the shadow of the moving object in the real image in the predicted frame, thereby reducing the shadow jitter problem existing in the predicted frame, reducing the difference between the predicted frame and the real image, and improving the user experience.
  • the image processing method shown in FIG3 and FIG4 mainly involves four parts: 1. obtaining the image data of the environmental shadow and the image data of the shadow of the moving object in the real image; 2. determining the position of the shadow of the moving object in the real image; 3. removing the shadow of the moving object generated according to the motion vector; 4. multiplexing the image data of the shadow of the moving object in the real image.
  • FIG5 shows a process of acquiring image data of environmental shadows and image data of shadows of moving objects in a real image, which may include the following steps:
  • the first FB may be GameFB, which may be used as the default FB of the application.
  • the image data obtained by executing the corresponding rendering according to the rendering instruction stream may be stored in GameFB. Determining to enable shadow rendering may refer to the above step S102, which will not be repeated here.
  • S302 Rendering environmental shadows to generate image data of environmental shadows.
  • the rendering instruction stream of the application program indicates that the environmental shadows should be rendered first, and then the shadow of the moving object should be superimposed on the environmental shadows to prevent the environmental shadows from overlapping the shadow of the moving object. Therefore, in this embodiment, the environmental shadows are rendered first, and the image data of the environmental shadows are stored in the first FB.
  • the second FB may be an AmbientFB
  • the third FB may be a HumanFB.
  • the purpose of creating the second FB and the third FB is to use the two FBs to store image data of environmental shadows and image data of shadows of moving objects respectively during the process of generating a real image, so as to intercept the image data of environmental shadows and image data of shadows of moving objects during the process of generating a real image.
  • step S303 is executed after determining to turn on shadow rendering, that is, step S303 and step S302 can be executed in parallel; in some examples, step S303 can be executed before step S302, which is not limited here.
  • S304 Copy the image data of the environmental shadow to the second FB, so as to realize the image processing of the environmental shadow by the second FB. Independent storage of data, so as to intercept the image data of the environment shadow without hindering the generation of the real image.
  • the glBlitFramebuffer() function is called to copy the image data of the environment shadow to the second FB.
  • S305 Change the FB bound to the application from the first FB to the third FB.
  • a possible implementation is to call the glBindFrameBuffer(Gl_DRAW_BUFFER, ID) function, where ID is the ID of the third FB, thereby changing the FB bound to the application from the first FB to the third FB.
  • the shadow of the moving object can be rendered. If the FB bound to the application is the first FB when rendering the shadow of the moving object, the image data of the shadow of the moving object is stored in the first FB, and the image data of the shadow of the moving object and the image data of the environment shadow are mixed and stored in the first FB, making it difficult to obtain the image data of the shadow of the moving object from the first FB.
  • the FB bound to the application is changed from the first FB to the third FB, and the image data of the shadow of the moving object generated when rendering the shadow of the moving object is stored in the third FB, thereby intercepting the image data of the shadow of the moving object in the process of generating a real image.
  • S306 Render the shadow of the moving object to generate image data of the shadow of the moving object.
  • the image data of the shadow of the moving object after the image data of the shadow of the moving object is generated, it can be copied to the first FB, which can also ensure that the image data of the environment shadow and the image data of the shadow of the moving object are stored in the first FB, but the efficiency of copying the image data is less than the efficiency of rendering the shadow of the moving object.
  • the first to third FBs may store pixel values of an R channel in the image data to reduce occupancy of the first to third FBs.
  • the image data of the environmental shadow is copied to the second FB, so that the image data of the environmental shadow is stored in the first FB and the second FB.
  • the third FB is bound to the application, which can store the image data of the shadow of the moving object in the third FB when rendering the shadow of the moving object, so as to store the image data of the environmental shadow in the second FB and the image data of the shadow of the moving object in the third FB, thereby completing the interception of the image data of the environmental shadow and the image data of the shadow of the moving object in the process of generating the real image.
  • the image data of the environmental shadow and the image data of the shadow of the moving object are stored in different FBs respectively, so as to prevent the two parts of the image data from overlapping and causing the problem of being difficult to distinguish.
  • the image data of the environmental shadow and the image data of the shadow of the moving object can be stored in the first FB defaulted by the application, so as to ensure the accuracy of the image data stored in the first FB and prevent the real image from lacking shadows.
  • the shadow rendering logic of the application is: when the shadow of the moving object is in the environmental shadow, the shadow of the moving object is not presented in the real image. Therefore, the image data of the environmental shadow and the image data of the shadow of the moving object are used to determine whether the shadow of the moving object is in the environmental shadow, so as to determine the position of the shadow of the moving object in the real image. As shown in FIG6 , it shows the process of determining the position of the shadow of the moving object in the real image, which may include the following steps:
  • the image data of the environment shadow can generate an environment shadow texture image
  • the image data of the shadow of the moving object can generate a moving object shadow texture image.
  • the pixel-by-pixel comparison calculation can be performed on the pixels with the same coordinates in the environment shadow texture image and the moving object shadow texture image.
  • step S403 determine whether the pixel is located in the ambient shadow, if not, execute step S404, if yes, execute step S405.
  • a pixel When a pixel is located in the shadow of a moving object but not in the shadow of the environment, it means that the pixel is a pixel in the shadow of the moving object, and the pixel is marked as being in the shadow area of the moving object.
  • a pixel When a pixel is not located in the shadow of a moving object, it means that the pixel is not a pixel in the shadow of the moving object, and the pixel is marked as not being in the shadow area of the moving object.
  • a pixel When a pixel is located in the shadow of a moving object and in the shadow of the environment, it means that the pixel is both a pixel in the shadow of the moving object and a pixel in the shadow of the environment. According to the principle that "when the shadow of a moving object is in the shadow of the environment, the shadow of the moving object is not presented in the real image", the pixel is marked as not being in the shadow area of the moving object.
  • a mask image is obtained, which represents the shadow area of the moving object in the real image.
  • the pixels in the shadow area of the moving object are marked as 1, and the pixels not in the shadow area of the moving object are marked as 0.
  • the area of the shadow of the moving object in the real image is determined by 1 and 0, thereby determining the position of the shadow of the moving object in the real image, that is, determining the position of the shadow of the moving object in the main image of the real image.
  • FIG. 7 shows a process of removing the shadow of a moving object generated according to a motion vector.
  • removing the shadow of the moving object in the first intermediate image is taken as an example, and the following steps may be included:
  • image distortion means moving the pixel according to the moving direction and moving speed of the pixel.
  • image distortion means moving the pixel according to the moving direction and moving speed of the pixel.
  • step S502 Determine whether the pixel is located in the shadow of the moving object according to the position of the shadow of the moving object in the second main image. If not, execute step S503; if yes, execute step S504.
  • each pixel point located in the shadow of the moving object in the second main image is still located in the shadow of the moving object after moving. Based on this principle, when the image is distorted, it is determined whether the pixel point is located in the shadow of the moving object in the second main image. If so, it is determined that the pixel point is located in the shadow of the moving object; otherwise, it is determined that the pixel point is not located in the shadow of the moving object.
  • the position of the shadow of the moving object in the second main image is known, and the position of the shadow of the moving object in the first intermediate image can be determined in advance according to the moving direction and moving speed in the motion vector that match the shadow of the moving object. After completing the movement of any pixel point in the second main image, it is determined whether the pixel point is located in the shadow of the moving object in the first intermediate image according to the coordinates of the pixel point in the first intermediate image and the position of the shadow of the moving object in the first intermediate image.
  • the pixel point in the first intermediate image are the same as the coordinates of a pixel point in the shadow of the moving object in the first intermediate image, it is determined that the pixel point is located in the shadow of the moving object in the first intermediate image; otherwise, it is determined that the pixel point is not located in the shadow of the moving object in the first intermediate image.
  • a mask image is generated based on the position of the shadow of the moving object in the first intermediate image, in which pixel points in the shadow area of the moving object are marked as 1, and pixel points that are not in the shadow area of the moving object are marked as 0.
  • the pixel value of the pixel point at the coordinates of the pixel point in the first intermediate image is searched from the mask image. If the pixel value of the pixel point at the coordinates is 1, it is determined that the pixel point is located in the shadow of the moving object; otherwise, it is determined that the pixel point is not located in the shadow of the moving object.
  • S503 Fill the pixel value of the pixel point with the pixel value of the pixel point in the second main image.
  • S505 After completing the image distortion of each pixel in the second main image, generate a first intermediate image with the shadow of the moving object removed, thereby obtaining a frame of the first intermediate image with the environment, the shadow of the environment, and the moving object, but with the shadow of the moving object removed.
  • the shadow of the moving object removed may be that the pixel value of each pixel in the shadow of the moving object in the first intermediate image is a preset value.
  • FIG8 is a process for multiplexing image data of shadows of moving objects in real images provided by the present application, which may include the following steps:
  • step S602. For each pixel point, determine whether the pixel point is located in the shadow of the moving object according to the position of the shadow of the moving object in the real image. If yes, execute step S603; if not, execute step S604.
  • whether the pixel point is located in the shadow of the moving object refers to whether the pixel point is located in the shadow of the moving object in the real image, specifically refers to the shadow of the moving object in the main image of the real image.
  • the position of the shadow of the moving object in the real image can be represented by the coordinates of each pixel point located in the shadow of the moving object. If the coordinates of the pixel point in the first intermediate image from which the shadow of the moving object is removed are the same as the coordinates of a pixel point located in the shadow of the moving object, then it is determined that the pixel point in the first intermediate image from which the shadow of the moving object is removed is located in the shadow of the moving object. Otherwise, it is determined that the pixel point in the first intermediate image from which the shadow of the moving object is removed is not located in the shadow of the moving object.
  • the position of the shadow of the moving object in the real image is represented by the mask image obtained in FIG. 6 above, specifically by a pixel point with a pixel value of 1 in the mask image. Then, when determining whether a pixel point in the first intermediate image for removing the shadow of the moving object is located in the shadow of the moving object, the coordinates of the pixel point are used to find the pixel value of the pixel point at the coordinates in the mask image. If the pixel value of the pixel point at the coordinates is 1, it is determined that the pixel point is located in the shadow of the moving object. Otherwise, it is determined that the pixel point is not located in the shadow of the moving object.
  • (1) in FIG9 shows a real image frame used to generate a prediction frame.
  • the position of the character's shadow in the real image is determined (such as the position circled by the irregular shape in (1) in FIG9).
  • the environment of the real image is distorted using the motion vector to obtain the intermediate image shown in (2) in FIG9, which removes the character's shadow and the character.
  • the character is distorted to obtain the intermediate image shown in (3) in FIG9, which removes the character's shadow and there is an area to be completed near the character.
  • the shadow is rendered for the intermediate image shown in (3) in FIG9 .
  • the image data of the shadow of the person in the real image shown in (1) in FIG9 is directly copied to the intermediate image shown in (3) in FIG9 , and the image shown in (4) in FIG9 is obtained.
  • Image completion processing is performed on each pixel in the area to be completed, and the image shown in (5) in FIG9 is obtained.
  • the shadow of the person in the image shown in (5) in FIG9 is obtained by reusing the shadow of the person shown in (1) in FIG9 , so the two frames of images have the same shadow.
  • the shadow of the object is the same, specifically, the image data and the position of the shadow of the person in the two frames of images are the same.
  • the present application also provides an electronic device, including: one or more processors and a memory, the memory is used to store computer program code, the computer program code includes computer instructions, when the one or more processors execute the computer instructions, the electronic device performs the above-mentioned image processing method.
  • the present application also provides a computer-readable storage medium for storing a computer program, which, when executed, is specifically used to implement the above-mentioned image processing method.
  • the present application also provides a computer program product comprising instructions, and when the computer program product is run on a computer or a processor, the computer or the processor executes the above-mentioned image processing method.

Landscapes

  • Engineering & Computer Science (AREA)
  • Physics & Mathematics (AREA)
  • General Physics & Mathematics (AREA)
  • Theoretical Computer Science (AREA)
  • Multimedia (AREA)
  • Computer Vision & Pattern Recognition (AREA)
  • Computer Graphics (AREA)
  • Processing Or Creating Images (AREA)

Abstract

本申请提供一种图像处理方法及装置，应用于电子设备中，方法包括：从第一真实图像中获取第一主图像以及从第二真实图像中获取第二主图像，第一真实图像和第二真实图像是电子设备渲染得到的图像；对第一主图像和第二主图像进行运动矢量计算，得到运动矢量；根据运动矢量，对目标主图像中的环境和目标主图像中的运动对象进行扭曲，生成预测帧中的环境和预测帧中的运动对象，目标主图像为第一主图像和第二主图像中的一帧主图像；在预测帧中复用运动对象的阴影在目标主图像中的图像数据，以生成预测帧中的运动对象的阴影。由此，减少了预测帧存在的阴影抖动问题，降低预测帧与真实图像之间的差异，提高用户体验。

Description

图像处理方法及装置
本申请要求于2022年11月21日提交中国国家知识产权局、申请号为202211454684.3、发明名称为“图像处理方法及装置”的中国专利申请的优先权，其全部内容通过引用结合在本申请中。
技术领域
本申请涉及图像处理技术领域，尤其涉及一种图像处理方法及装置。
背景技术
预测帧技术是图像渲染中的一个主流技术，电子设备使用预测帧技术可以降低图像渲染的开销。但是为了增强用户的沉浸体验感，电子设备添加了阴影渲染，预测帧技术使得预测帧存在阴影抖动问题，增大预测帧与真实图像之间的差异，极大的降低了用户体验，真实图像是电子设备根据渲染逻辑渲染生成的图像，且真实图像用于生成预测帧。
发明内容
本申请提供了一种图像处理方法及装置，目的在于解决预测帧存在阴影抖动问题，以降低预测帧与真实图像之间的差异，提高用户体验。为了实现上述目的，本申请提供了以下技术方案：
第一方面，本申请提供一种图像处理方法，应用于电子设备中，方法包括：从第一真实图像中获取第一主图像以及从第二真实图像中获取第二主图像，第一真实图像和第二真实图像是电子设备渲染得到的图像；对第一主图像和第二主图像进行运动矢量计算，得到运动矢量；根据运动矢量，对目标主图像中的环境和目标主图像中的运动对象进行扭曲，生成预测帧中的环境和预测帧中的运动对象，目标主图像为第一主图像和第二主图像中的一帧主图像；在预测帧中复用运动对象的阴影在目标主图像中的图像数据，以生成预测帧中的运动对象的阴影。由此，预测帧中的运动对象的阴影可以从预测帧中的环境中抽离出来，使得预测帧中的运动对象的阴影能够独立生成，且预测帧中的运动对象的阴影不再依赖运动矢量，而是通过在预测帧中复用运动对象的阴影在目标主图像中的图像数据生成，减少了预测帧存在的阴影抖动问题，降低预测帧与真实图像之间的差异，提高用户体验。
在一些示例中，根据运动矢量，对目标主图像中的环境和目标主图像中的运动对象进行扭曲，生成预测帧中的环境和预测帧中的运动对象，在预测帧中复用运动对象的阴影在目标主图像中的图像数据，以生成预测帧中的运动对象的阴影包括：根据运动矢量，对目标主图像中的环境和目标主图像中的运动对象进行扭曲，生成第一中间图像，第一中间图像包括环境和运动对象；在第一中间图像中复用运动对象的阴影在目标主图像中的图像数据，使得第一中间图像包括环境、运动对象和运动对象的阴影。在得到包括环境、运动对象和运动对象的阴影的第一中间图像后，对第一中间图像进行图像补全，得到第二中间图像；然后对第二中间图像进行后期处理，将目标主图像对应的用户界面图像和处理后的第二中间图像进行融合，得到预测帧。因为在得到第一中间图像后，运动对象和运动对象的阴影没有变化，所以第一中间图像中的运动对象可以是预测帧中的运动对象，第一中间图 像中的运动对象的阴影可以是预测帧中的运动对象的阴影。对于第一中间图像中的环境来说在图像补全过程中被补全，因此预测帧包括第一中间图像中已经补全的环境，即第二中间图像中的环境。
在一种可能的实现方式中，在预测帧中复用运动对象的阴影在目标主图像中的图像数据包括：对预测帧中的各像素点，根据运动对象的阴影在目标主图像中的位置，确定像素点是否位于目标主图像的运动对象的阴影中；在确定像素点位于目标主图像的运动对象的阴影中，复用像素点在目标主图像中的像素值，像素点在目标主图像中的像素值为像素点在目标主图像中的图像数据；在确定像素点没有位于目标主图像的运动对象的阴影中，保持像素点的像素值不变。由此，对预测帧中位于目标主图像的运动对象的阴影中的像素点，复用了该像素点在目标主图像中的像素值，从而既复用了运动对象的阴影在目标主图像中的位置，又复用了运动对象的阴影在目标主图像中的图像数据，以将目标主图像中的运动对象的阴影融合到预测帧中，避免了继续使用环境所用的运动矢量生成预测帧中的运动对象的阴影，减少了预测帧存在的阴影抖动问题，降低预测帧与真实图像之间的差异，提高用户体验。因为预测帧复用了运动对象的阴影在目标主图像中的位置，所以运动对象的阴影在预测测帧中的位置与其在目标主图像中的位置相同，在复用了运动对象的阴影在目标主图像中的图像数据的情况下，运动对象的阴影在预测帧和目标主图像中的形状相同、颜色也相同、亮度也相同。
在一种可能的实现方式中，运动矢量用于记录目标主图像中各像素点的移动速度和移动方向，根据运动矢量，对目标主图像中的环境和目标主图像中的运动对象进行扭曲，生成预测帧中的环境和预测帧中的运动对象包括：根据目标主图像中各像素点的移动速度和移动方向，对目标主图像中的各像素点进行移动；在对像素点进行移动过程中，根据运动对象的阴影在目标主图像中的位置，确定像素点是否位于预测帧的运动对象的阴影中；在确定像素点位于预测帧的运动对象的阴影中，将像素点的像素值设置为预设值；在确定像素点没有位于预测帧的运动对象的阴影中，将像素点的像素值设置为像素点在目标主图像中的像素值。由此，以预设值来区分预测帧的运动对象的阴影与预测帧中的其他对象，从而通过预设值抠除预测帧中的运动对象的阴影并能够标注出运动对象的阴影的位置，以为预测帧预留出运动对象的阴影，使得预测帧可以复用运动对象的阴影在目标主图像中的图像数据。
在一种可能的实现方式中，预设值包括预设透明度，预设透明度与目标主图像中其他像素点的透明度不同，通过预设透明度进行区分。
在一种可能的实现方式中，运动对象的阴影在目标主图像中的位置的确定过程包括：将环境阴影纹理图像和运动对象阴影纹理图像中位置相同的各待比对像素点进行比对，环境阴影纹理图像根据目标主图像中环境阴影的图像数据生成，运动对象阴影纹理图像根据目标主图像中运动对象的阴影的图像数据生成；在确定待比对像素点处于运动对象的阴影区域但没有处于环境阴影中，标记待比对像素点处于运动对象的阴影区域中；在确定待比对像素点处于运动对象的阴影区域且处于环境阴影中，标记待比对像素点没有处于运动对象的阴影区域中；根据标记为处于运动对象的阴影区域中的待比对像素点，确定运动对象的阴影在目标主图像中的位置，运动对象的阴影在目标主图像中的图像数据包括位于位置 中的待比对像素点在目标主图像中的像素值。在一些示例中，运动对象的阴影在目标主图像中的图像数据与目标主图像中运动对象的阴影的图像数据相同。通过环境阴影纹理图像和运动对象阴影纹理图像，能够按照“当运动对象的阴影位于环境阴影中，在目标主图像中不呈现运动对象的阴影”的原则，确定出运动对象的阴影在目标主图像中的位置，为确定运动对象的阴影在预测帧中的位置做准备，从而能够复用运动对象的阴影的图像数据。
在一种可能的实现方式中，若运动对象阴影纹理图像中待比对像素点的像素值大于环境阴影纹理图像中待比对像素点的像素值，确定待比对像素点处于运动对象的阴影区域但没有处于环境阴影中；若运动对象阴影纹理图像中待比对像素点的像素值小于或等于环境阴影纹理图像中待比对像素点的像素值，确定待比对像素点处于运动对象的阴影区域且处于环境阴影中。
在一种可能的实现方式中，待比对像素点的像素值包括待比对像素点的R通道值、G通道值和B通道值中的至少一个。在一些示例中，待比对像素点的像素值包括R通道值、G通道值和B通道值中的一个，可以节省存储空间；在一些示例中，待比对对像素点的像素值包括R通道值、G通道值和B通道值中的两个或三个，那么在比对像素值时可以是先对待比对像素点的像素值进行融合处理(如加权平均)，得到一个数值，然后在对数值进行比对，从而综合考虑不同通道值的影响。
在一种可能的实现方式中，在从第一真实图像中获取第一主图像以及从第二真实图像中获取第二主图像之前，方法还包括：在渲染目标主图像对应的真实图像时，得到环境阴影的图像数据和运动对象的阴影的图像数据，环境阴影的图像数据和运动对象的阴影的图像数据用于确定运动对象的阴影在目标主图像中的位置，运动对象的阴影在目标主图像中的位置和运动对象的阴影的图像数据用于得到待复用的运动对象的阴影在目标主图像中的图像数据。由此，在渲染目标主图像对应的真实图像过程中，完成环境阴影的图像数据和运动对象的阴影的图像数据的拦截，为复用做准备。
在一种可能的实现方式中，在渲染目标主图像对应的真实图像时，得到环境阴影的图像数据和运动对象的阴影的图像数据包括：渲染真实图像的环境阴影，以生成环境阴影的图像数据，环境阴影的图像数据存储到第一帧缓冲中；创建第二帧缓冲和第三帧缓冲；将第一帧缓冲中存储的环境阴影的图像数据复制到第二帧缓冲中；将真实图像绑定的帧缓冲从第一帧缓冲改为第三帧缓冲；渲染真实图像的运动对象的阴影，以生成运动对象的阴影的图像数据，运动对象的阴影的图像数据存储到第三帧缓冲中。由此环境阴影的图像数据和运动对象的阴影的图像数据是分别存储在不同的帧缓冲中，防止这两部分图像数据重叠，造成难以区分的问题，从而方便使用环境阴影的图像数据和运动对象的阴影的图像数据。
在一种可能的实现方式中，在渲染目标主图像对应的真实图像时，得到环境阴影的图像数据和运动对象的阴影的图像数据还包括：在渲染完真实图像的运动对象的阴影后，将真实图像绑定的帧缓冲从第三帧缓冲改为第一帧缓冲；再次渲染真实图像的运动对象的阴影，以生成运动对象的阴影的图像数据，运动对象的阴影的图像数据存储到第一帧缓冲中。在实现拦截环境阴影的图像数据和运动对象的阴影的图像数据的同时，能够在真实图像绑定的第一帧缓冲中存储有环境阴影的图像数据和运动对象的阴影的图像数据，保证第一帧缓冲中存储的图像数据的准确度，防止真实图像(主要是目标主图像)缺少阴影。
在一种可能的实现方式中，将第一帧缓冲中存储的环境阴影的图像数据复制到第二帧缓冲中包括：调用glBlitFramebuffer函数将第一帧缓冲中存储的环境阴影的图像数据复制到第二帧缓冲中。
在一种可能的实现方式中，将真实图像绑定的帧缓冲从第一帧缓冲改为第三帧缓冲包括：调用glBindFrameBuffer函数将真实图像绑定的帧缓冲从第一帧缓冲改为第三帧缓冲。
在一种可能的实现方式中，方法还包括：在生成预测帧中的运动对象的阴影后，对预测帧进行图像补全。
第二方面，本申请提供一种电子设备，包括：一个或多个处理器以及存储器；存储器用于存储计算机程序代码，计算机程序代码包括计算机指令，当一个或多个处理器执行计算机指令时，电子设备执行第一方面所述的图像处理方法。
第三方面，本申请提供一种计算机可读存储介质，其用于存储计算机程序，计算机程序被执行时，具体用于实现如第一方面所述的图像处理方法。
本申请在上述各方面提供的实现方式的基础上，还可以进行进一步组合以提供更多实现方式。
附图说明
图1为本申请提供的生成预测帧的示意图；
图2为本申请提供的电子设备的硬件结构图；
图3为本申请提供的生成真实图像的流程图；
图4为本申请提供的生成预测帧的流程图；
图5为本申请提供的获取真实图像中环境阴影的图像数据和运动对象的阴影的图像数据的流程图
图6为本申请提供的确定运动对象的阴影在真实图像中的位置的流程图；
图7为本申请提供的抠除根据运动矢量生成的运动对象的阴影的流程图；
图8为本申请提供的复用真实图像中运动对象的阴影的图像数据的流程图；
图9为本申请提供的复用阴影的示意图。
具体实施方式
下面将结合本申请实施例中的附图，对本申请实施例中的技术方案进行清楚、完整地描述。以下实施例中所使用的术语只是为了描述特定实施例的目的，而并非旨在作为对本申请的限制。如在本申请的说明书和所附权利要求书中所使用的那样，单数表达形式“一个”、“一种”、“所述”、“上述”、“该”和“这一”旨在也包括例如“一个或多个”这种表达形式，除非其上下文中明确地有相反指示。还应当理解，在本申请实施例中，“一个或多个”是指一个、两个或两个以上；“和/或”，描述关联对象的关联关系，表示可以存在三种关系；例如，A和/或B，可以表示：单独存在A，同时存在A和B，单独存在B的情况，其中A、B可以是单数或者复数。字符“/”一般表示前后关联对象是一种“或”的关系。
在本说明书中描述的参考“一个实施例”或“一些实施例”等意味着在本申请的一个或多个实施例中包括结合该实施例描述的特定特征、结构或特点。由此，在本说明书中的不同之处出现的语句“在一个实施例中”、“在一些实施例中”、“在其他一些实施例中”、“在另外一些实施例中”等不是必然都参考相同的实施例，而是意味着“一个或多个但不是所有的实 施例”，除非是以其他方式另外特别强调。术语“包括”、“包含”、“具有”及它们的变形都意味着“包括但不限于”，除非是以其他方式另外特别强调。
本申请实施例涉及的多个，是指大于或等于两个。需要说明的是，在本申请实施例的描述中，“第一”、“第二”等词汇，仅用于区分描述的目的，而不能理解为指示或暗示相对重要性，也不能理解为指示或暗示顺序。
在用户使用电子设备的过程中，电子设备通过显示屏可以向用户展示多媒体流文件，以便于向用户提供丰富的视觉体验。以多媒体流为视频流为例。该视频流可以包括多帧图像。示例性的，该视频流可以包括N帧图像，如第1帧，第2帧，……第N帧。电子设备可以在展示该视频流时，分别在显示屏上显示第1帧，第2帧，……第N帧。想要播放视频流的应用程序，可以向电子设备发出针对不同帧图像的渲染命令。电子设备可以根据这些渲染命令，对各帧图像进行渲染，得到各帧图像的渲染结果(如图像数据)，并基于渲染结果进行显示。
随着应用程序的迭代更新，电子设备根据渲染命令进行图像渲染的流程越来越复杂，导致电子设备的资源负载也越来越大；另一方面，电子设备的电池续航能力有限，在保证图像的画质不受影响的情况下，减少不必要的图像渲染的开销，成为一大难题。针对此，研究人员研发了预测帧技术，预测帧技术可以是基于两帧图像生成一帧图像，即基于两帧图像生成一帧预测帧，为了区分，生成预测帧的图像称为真实图像。
图1是生成预测帧的示意图，生成预测帧的流程是：电子设备对第一真实图像和第二真实图像进行用户界面(User Interface，UI)分离，得到第一真实图像的第一主图像和第一真实图像的第一UI图像、第二真实图像的第二主图像和第二真实图像的第二UI图像，第一真实图像是第二真实图像的前一帧图像，UI图像包括真实图像中UI界面，主图像包括真实图像中环境、环境阴影、运动对象、运动对象的阴影等对象。电子设备根据第一主图像和第二主图像，进行运动矢量(Motion Vector，MV)计算，得到运动矢量，运动矢量指示出第二主图像中各对象的移动方向和移动速度；根据运动矢量对第二主图像进行图像扭曲(ImageWarp)，得到第一中间图像，其中ImageWarp是利用运动矢量对第二主图像中的各对象进行移动；由于第二主图像中各对象的移动速度和移动方向不同，第一中间图像出现图像数据缺失的情况，因此在得到第一中间图像后，对第一中间图像进行图像补全(HoleFilling)，得到第二中间图像；对第二中间图像和第二UI图像进行融合(即第二UI图像贴合到第二中间图像上)，生成预测帧，实现根据第一真实图像和第二真实图像生成预测帧。
在一些示例中，预测帧可以是第一真实图像和第二真实图像的中间帧；在一些示例中，预测帧可以是第二真实图像的下一帧图像。此外，电子设备在得到运动矢量后，可以根据运动矢量对第一主图像进行ImageWarp，以得到第一中间图像，相对应的，在得到第二中间图像后，对第二中间图像和第一UI图像进行融合，以生成预测帧。
为了增强用户的沉浸体验感，电子设备可以在图像渲染中添加阴影渲染，即对于第一真实图像和第二真实图像中的运动对象，电子设备为运动对象渲染了阴影，阴影随着运动对象的运动而运动。例如针对游戏应用程序中的人物，为人物渲染了阴影，在人物运动时人物的阴影随着人物运动，那么生成人物的预测帧时，预测帧也是具有阴影的。但是一帧 图像除了包括运动对象、运动对象的阴影之外，还包括环境，环境指向运动对象所处场景，电子设备对该环境也渲染了阴影，称为环境阴影。电子设备生成预测帧时，运动对象的阴影按照环境的移动方向和移动速度进行移动，存在错误移动运动对象的阴影的现象，导致运动对象运动过程中，预测帧存在阴影抖动问题，增大预测帧与真实图像之间的差异，极大的降低了用户体验。
针对这一问题，本申请提供一种图像处理方法，在生成预测帧过程中，复用真实图像中运动对象的阴影的图像数据到预测帧中，以生成预测帧中运动对象的阴影，预测帧中的运动对象、环境和环境阴影通过运动矢量、一帧真实图像的主图像和该真实图像的UI图像生成。如复用第一真实图像或第二真实图像中运动对象的阴影的图像数据。
为了能够复用真实图像中运动对象的阴影的图像数据，在生成预测帧过程中，抠除根据运动矢量生成的运动对象的阴影。并且，电子设备渲染真实图像时，获取真实图像中运动对象的阴影的图像数据以及运动对象的阴影在真实图像中的位置，其中运动对象的阴影在真实图像中的位置是为了确定出运动对象的阴影在预测帧中的位置，真实图像中运动对象的阴影的图像数据是为了在预测帧中复用。由此，预测帧中运动对象的阴影不再依赖环境的移动方向和移动速度，而是通过在预测帧中复用真实图像中运动对象的阴影的图像数据生成，减少了预测帧存在的阴影抖动问题，降低预测帧与真实图像之间的差异，提高用户体验。真实图像的主图像包括运动对象的阴影，复用真实图像中运动对象的阴影的图像数据可以是指复用该真实图像中主图像中运动对象的阴影的图像数据，运动对象的阴影在真实图像中的位置可以是指运动对象的阴影在主图像中的位置。其中复用的主图像中运动对象的阴影的图像数据也可以称为运动对象的阴影在主图像中的图像数据。
应用程序的阴影渲染逻辑是：当运动对象的阴影位于环境阴影中，在真实图像中不呈现运动对象的阴影，因此在确定运动对象的阴影在真实图像中的位置时，存在去掉既位于于运动对象的阴影中又位于环境阴影中的情况，由此，在一些示例中，运动对象的阴影在主图像中的图像数据是位于运动对象的阴影中但没有位于环境阴影中的图像数据，即主图像中运动对象的阴影的部分图像数据；例如在下述步骤S306获取到运动对象的阴影的全部图像数据，运动对象的阴影在主图像中的图像数据是步骤S306获取到的全部图像数据中的部分图像数据。在一些示例中，运动对象的阴影在主图像中的图像数据与步骤S306获取到的全部图像数据相同。
以下结合附图对本申请实施例提供的方案进行详细说明。
需要说明的是，本申请实施例提供的图像处理方法，可以应用在用户的电子设备中。该电子设备可以是能够提供网络接入能够的设备。比如，该电子设备可以是手机、平板电脑、桌面型、膝上型、笔记本电脑、超级移动个人计算机(Ultra-mobile Personal Computer，UMPC)、手持计算机、上网本、个人数字助理(Personal Digital Assistant，PDA)、可穿戴电子设备、智能手表等设备。本申请对电子设备的具体形式不做特殊限定。在一些实施例中，该电子设备可以具有显示功能。比如，电子设备可以根据第一真实图像和第二真实图像，生成预测帧，显示该预测帧，以及显示第一真实图像和第二真实图像。
如图2所示，该电子设备可以包括：处理器，外部存储器接口，内部存储器，通用串行总线(Universal Serial Bus，USB)接口，充电管理模块，电源管理模块，电池，天线1， 天线2，移动通信模块，无线通信模块，传感器模块，按键，马达，指示器，摄像头，显示屏，以及用户标识模块(Subscriber Identity Module，SIM)卡接口等。其中音频模块可以包括扬声器，受话器，麦克风，耳机接口等，传感器模块可以包括压力传感器，陀螺仪传感器，气压传感器，磁传感器，加速度传感器，距离传感器，接近光传感器，指纹传感器，温度传感器，触摸传感器，环境光传感器，骨传导传感器等。
可以理解的是，本申请实施例示意的结构并不构成对电子设备的具体限定。在另一些实施例中，电子设备可以包括比图示更多或更少的部件，或者组合某些部件，或者拆分某些部件，或者不同的部件布置。图示的部件可以以硬件，软件或软件和硬件的组合实现。
其中，处理器可以包括一个或多个处理单元，例如：处理器可以包括应用处理器(Application Processor，AP)，调制解调处理器，图形处理器(Graphics Processing Unit，GPU)，图像信号处理器(Image Signal Processor，ISP)，控制器，视频编解码器，数字信号处理器(Digital Signal Processor，DSP)，基带处理器，和/或神经网络处理器(Neural-network Processing Unit，NPU)等。其中，不同的处理单元可以是独立的器件，也可以集成在一个或多个处理器中。处理器是电子设备的神经中枢和指挥中心，控制器可以根据指令操作码和时序信号，产生操作控制信号，完成取指令和执行指令的控制。
外部存储器接口可以用于连接外部存储卡，例如Micro SD卡，实现扩展电子设备的存储能力。外部存储卡通过外部存储器接口与处理器通信，实现数据存储功能。内部存储器可以用于存储计算机可执行程序代码，所述可执行程序代码包括指令。处理器通过运行存储在内部存储器的指令，从而执行电子设备的各种功能应用以及数据处理。例如，在本申请实施例中，处理器通过运行内部存储器中存储的指令，使得电子设备执行本申请实施例提供的图像处理方法。
电子设备通过GPU，显示屏，以及应用处理器等实现显示功能。GPU为图像处理的微处理器，连接显示屏和应用处理器。GPU用于执行数学和几何计算，用于图像渲染。处理器可包括一个或多个GPU，其执行程序指令以生成或改变显示信息。
显示屏用于显示图像，视频等。显示屏包括显示面板。显示面板可以采用液晶显示屏(liquid crystal display，LCD)，有机发光二极管(organic light-emitting diode，OLED)，有源矩阵有机发光二极体或主动矩阵有机发光二极体(active-matrix organic light emitting diode，AMOLED)，柔性发光二极管(flex light-emitting diode，FLED)，Miniled，MicroLed，Micro-oled，量子点发光二极管(quantum dot light emitting diodes，QLED)等。在一些实施例中，电子设备可以包括1个或N个显示屏，N为大于1的正整数。
电子设备的显示屏上可以显示一系列图形用户界面(graphical user interface，GUI)，这些GUI都是该电子设备的主屏幕。一般来说，电子设备的显示屏的尺寸是固定的，只能在该电子设备的显示屏中显示有限的控件。控件是一种GUI元素，它是一种软件组件，包含在应用程序中，控制着该应用程序处理的所有数据以及关于这些数据的交互操作，用户可以通过直接操作(direct manipulation)来与控件交互，从而对应用程序的有关信息进行读取或者编辑。一般而言，控件可以包括图标、按钮、菜单、选项卡、文本框、对话框、状态栏、导航栏、Widget等可视的界面元素。例如，在本申请实施例中，显示屏可以显示包含特定动作的视频文件等等。
在一些实施例中，处理器包括CPU和GPU，CPU可以根据应用程序的渲染逻辑生成渲染指令流，向GPU下发渲染指令流，以便于GPU根据渲染指令流执行对应的渲染。其中，CPU可以向GPU逐条发送渲染指令流的渲染指令。在一些示例中，CPU可以接收到GPU的响应后再次向GPU发送渲染指令；在一些示例中，CPU不需要等待GPU的响应，在完成一条渲染指令的发送后继续发送下一条渲染指令，具体看CPU对渲染指令的设置。
在本申请实施例中，GPU在执行真实图像的渲染过程中，GPU可以获取到真实图像中运动对象的阴影的图像数据，以在生成预测帧时复用这部分图像数据。如图3所示，其示出了本申请实施例提供的图像处理方法中生成真实图像的流程，在生成真实图像过程中获取真实图像中运动对象的阴影的图像数据，可以包括以下步骤：
S101、接收渲染指令流。渲染指令流包括至少一条渲染指令，通过渲染指令指示GPU执行对应的渲染，如指示GPU渲染真实图像中的运动对象、指示GPU渲染真实图像中的环境等等。
S102、根据渲染指令流，确定是否开启阴影渲染，如果是，执行步骤S103；如果否，执行步骤S107。
在本申请实施例中，预先对应用程序进行离线渲染分析，比对应用程序开启阴影渲染和关闭阴影渲染时的渲染指令的调用规律和渲染指令的特性，得到该应用程序用于识别开启阴影渲染的判断条件。在接收到渲染指令流后，判断渲染指令流携带的信息是否满足判断条件，如果满足，确定开启了阴影渲染；如果不满足，确定关闭了阴影渲染。
例如，在一些示例中，应用程序开启阴影渲染时，应用程序调用特定程序、特定资源和特定的渲染指令中的至少一种，由此，判断条件可以记录应用程序开启阴影渲染时调用的信息。如果应用程序开启阴影渲染时调用特定程序，判断条件可以记录特定程序的标签(label)，以通过该标签识别特定程序，例如特定程序的标签可以是特定程序的程序名称；如果应用程序开启阴影渲染时调用特定资源，判断条件可以记录特定资源的资源标签，以通过资源标签识别特定资源，例如资源标签可以是资源名称等；如果应用程序开启阴影渲染时调用特定指令，判断条件可以记录该特定指令。
例如游戏应用程序开启阴影渲染时调用glClearColor(1.0,1.0,1.0,0.0)，因此判断条件是glClearColor(1.0,1.0,1.0,0.0)，既要求调用glClearColor(R,G,B,A)且要求(RGBA＝＝1.0,1.0,1.0,0.0)，由此渲染指令流中存在该条渲染指令-glClearColor(1.0,1.0,1.0,0.0)时，确定开启了阴影渲染，其中R、G、B、A分别是红色、绿色、蓝色和透明度。
S103、根据渲染指令流中与环境阴影渲染相关的渲染指令(ScreenSpaceShadow Ambient Shadow)，渲染环境阴影(Ambient Shadow)。渲染环境阴影是为了生成环境阴影的图像数据。
S104、在渲染环境阴影时拦截环境阴影的渲染，以获取到环境阴影的图像数据。
S105、根据渲染指令流中与运动对象的阴影渲染相关的渲染指令(ScreenSpaceShadow Human Shadow)，渲染运动对象的阴影(Human Shadow)。渲染运动对象的阴影是为了生成运动对象的阴影的图像数据。
S106、在渲染运动对象的阴影时拦截运动对象的阴影的渲染，以获取到运动对象的阴影的图像数据。
S107、根据渲染指令流执行对应的渲染，以得到真实图像的全部图像数据，通过步骤S103至步骤S106完成了环境阴影和运动对象的阴影的渲染，一帧真实图像还可以包括运动对象、环境和UI界面等，由此，在一些示例中，此处的渲染主要是对图像中除环境阴影和运动对象的阴影之外的渲染，如对运动对象、环境、UI界面等的渲染。在一些示例中，运动对象和环境的渲染可以是阴影渲染之前，那么步骤S107的渲染可以是对图像中除环境阴影、运动对象的阴影、运动对象和环境之外的渲染，如对UI界面的渲染。
在一些示例中，在执行步骤S104获取环境阴影的图像数据过程中，可以执行步骤S105渲染运动对象的阴影。在一些示例中，在执行步骤S106获取运动对象的阴影的图像数据过程中，可以执行步骤S107根据渲染指令流执行对应的渲染，以完成真实图像的渲染。
S108、根据真实图像的全部图像数据，输出真实图像。在得到真实图像的全部图像数据后，根据真实图像的全部图像数据，生成真实图像，并显示真实图像。
在图3所示流程中，在生成真实图像的过程中，获取到环境阴影的图像数据和运动对象的阴影的图像数据。图4示出了生成预测帧的流程，在生成预测帧的过程中，使用了环境阴影的图像数据和运动对象的阴影的图像数据，可以包括以下步骤：
S201、输入第一真实图像和第二真实图像。
S202、根据第一真实图像的第一主图像和第二真实图像的第二主图像，进行MV计算，得到运动矢量。即对第一主图像和第二主图像进行MV计算，得到运动矢量，运动矢量指示第二主图像中各对象的移动方向和移动速度。
S203、确定是否开启阴影渲染，如果是，执行步骤S204，如果否，执行步骤S208。其中，确定是否开启阴影渲染可以参见上述步骤S102，此处不再赘述。
S204、根据环境阴影的图像数据和运动对象的阴影的图像数据，确定真实图像中运动对象的阴影所在区域。
S205、根据运动矢量对第二主图像进行图像扭曲，得到第一中间图像，并抠除第一中间图像中运动对象的阴影。
其中，根据运动矢量对第二主图像进行图像扭曲主要是利用运动矢量对第二主图像中的各对象进行移动，具体是对第二主图像中的任一对象，根据运动矢量中该对象的移动方向和移动速度，移动该对象。第二主图像包括环境、环境阴影、运动对象和运动对象的阴影，在根据运动矢量对第二主图像中的各对象进行移动时，先对第二主图像进行运动对象和环境的分离，此次分离将环境和运动对象的阴影与运动对象分离开，即将第二主图像分离成两部分，第一部分包括环境、环境阴影和运动对象的阴影，第二部分包括运动对象。
然后，利用运动矢量中第一部分中各对象的移动方向和移动速度，对第一部分中各对象进行移动；再利用运动矢量中第二部分的运动对象的移动方向和移动速度，对运动对象进行移动。在一些示例中，在移动运动对象的阴影时，禁止复用真实图像中运动对象的阴影的图像数据，通过此方式实现了抠除第一中间图像中运动对象的阴影，并能够在生成第一中间图像的过程中完成对运动对象的阴影的抠除。在一些示例中，在生成第一中间图像后，再将第一中间图像中运动对象的阴影抠除。无论哪种方式，步骤S205得到的第一中间图像是一帧有环境、环境阴影、运动对象，但是抠除运动对象的阴影的图像。
抠除运动对象的阴影的目的是在第一中间图像中标记出运动对象的阴影的位置，标记 出的位置在生成第一中间图像过程中没有使用真实图像中运动对象的阴影的图像数据，以在标记出的位置预留了一个像素值待补全的待补全区域。第一中间图像标记的运动对象的阴影的位置是根据运动矢量移动真实图像中运动对象的阴影得到，因此第一中间图像标记的运动对象的阴影的位置与其在真实图像中的位置不同。
S206、在第一中间图像中复用第二主图像中运动对象的阴影的图像数据。
在一些示例中，复用第二主图像中运动对象的阴影的图像数据不单单是复用图像数据还复用了运动对象的阴影在第二主图像中的位置，使得第二主图像的运动对象的阴影贴到了第一中间图像中。但是第一中间图像标记的运动对象的阴影的位置与其在第二主图像中的位置不同，因此在复用运动对象的阴影的图像数据后，第一中间图像标记的运动对象的阴影的位置仍存在像素值待补全的待补全区域，该待补全区域中的各像素点的像素值通过图像补全处理得到。
S207、根据运动矢量对第二主图像进行图像扭曲，得到第一中间图像。
S208、对第一中间图像进行图像补全(HoleFilling)处理，得到第二中间图像。
由于第一中间图像是对第二主图像中的各对象移动得到，第二主图像中各对象的移动速度和移动方向可能不同，使得第一中间图像出现图像数据缺失的情况，图像数据缺失的区域可以称为待补全区域。待补全区域中各像素点的像素值通过图像补全处理得到。
在一种实现方式中，可以采用深度学习网络将待补全区域中各像素点的像素值补全。在另一种实现方式中，可以采用传统的图像修复操作将补全区域中各像素点的像素值补全。在另一种实现方式中，利用第一中间图像的掩膜(Mask)图像对待补全区域中各像素点的像素值进行补全，用于确定与待补像素点对应的一些静态像素点，这些静态像素点的像素值用于确定待补像素点的像素值。
在掩膜图像中，包括三类像素点，一类是待补像素点，一类是静态像素点，一类是动态像素点。其中，待补像素点指的是掩膜图像的待补全区域(与第一中间图像相同)中需要进行像素值补全的像素点，一般运动对象和运动对象的阴影分别对应一个待补全区域，待补全区域可以是一个空白区域，那么运动对象附近空白区域中的像素点为待补像素点，运动对象的阴影附近空白区域中的像素点为待补像素点；静态像素点指的是与掩膜图像中处于静止状态的环境所对应的像素点，其像素值可以用于确定待补像素点的像素值；动态像素点指的是与掩膜图像中处于运动状态的对象(或称运动对象)所对应的像素点，其像素值不可以用于确定待补像素点的像素值。
也即，静态像素点是用于对待补像素点进行补全的有效参考点，动态像素点是用于对待补像素点进行补全的无效参考点。在掩膜图像中，如上三类像素点可以通过灰度值进行区分，三类像素点的灰度值不同。待补像素点的灰度值为第一灰度值，静态像素点的灰度值为第二灰度值，动态像素点的灰度值为第三灰度值。其中，第一灰度值、第二灰度值和第三灰度值互不相等。示例性的，静态像素点的灰度值为0，动态像素点的灰度值为1，待补像素点的灰度值为0.5。
针对掩膜图像中的每个待补像素点，电子设备在掩膜图像中确定与待补像素点对应的一组有效静态像素点，并基于这组有效静态像素点在第一中间图像中的像素值确定待补像素点的像素值。一种可能的实现方式是，将一组有效静态像素点在第一中间图像中的像素 值加权平均后作为待补像素点的像素值。电子设备根据各个待补像素点的像素值对第一中间图像进行像素值填充，得到第二中间图像。
其中，与某个待补像素点对应的一组有效静态像素点，与这个待补像素点的像素间距相等。而且，与其他静态像素点(可以称之为无效静态像素点)相比，这组有效静态像素点与这个待补像素的像素间距最小。
S209、对第二中间图像进行后期处理(PostProcess)。在一些示例中，后期处理主要是对第二中间图像进行图像滤波。
S210、对第二真实图像的第二UI图像和处理后的第二中间图像进行融合，得到预测帧。
从上述图3和图4可知，在生成真实图像的过程中，获取到环境阴影的图像数据和运动对象的阴影的图像数据；在生成预测帧的过程中，利用环境阴影的图像数据和运动对象的阴影的图像数据，确定出运动对象的阴影在真实图像中的位置，并利用运动对象的阴影在真实图像中的位置，抠除根据运动矢量生成的运动对象的阴影，以标注出根据运动矢量生成的运动对象的阴影在图像中的位置，然后复用真实图像中运动对象的阴影的图像数据，以生成预测帧中运动对象的阴影，由此，预测帧中运动对象的阴影能够从预测帧的环境中抽离出来，使得预测帧中运动对象的阴影不再依赖运动矢量生成，而是通过在预测帧中复用真实图像中运动对象的阴影的图像数据生成，减少了预测帧存在的阴影抖动问题，降低预测帧与真实图像之间的差异，提高用户体验。
并且，在解决预测帧存在阴影抖动问题的场景下，上述图3和图4所示图像处理方法主要涉及四部分内容：一、获取到真实图像中环境阴影的图像数据和运动对象的阴影的图像数据；二、确定出运动对象的阴影在真实图像中的位置；三、抠除根据运动矢量生成的运动对象的阴影；四、复用真实图像中运动对象的阴影的图像数据。下面结合附图对这四部分内容进行说明。
图5示出了获取真实图像中环境阴影的图像数据和运动对象的阴影的图像数据的流程，可以包括以下步骤：
S301、确定开启阴影渲染，并确定当前应用程序绑定的帧缓冲(Frame Buffer，FB)为第一FB。其中，第一FB可以是GameFB，第一FB可以作为应用程序的默认FB，根据渲染指令流执行对应渲染所得到的图像数据，可以存储到GameFB中。确定开启阴影渲染可以参见上述步骤S102，此处不再赘述。
S302、渲染环境阴影，以生成环境阴影的图像数据。应用程序的渲染指令流指示先渲染环境阴影，然后在环境阴影上叠加运动对象的阴影，以防止环境阴影重叠到运动对象的阴影上，因此本实施例首先渲染环境阴影，环境阴影的图像数据存储到第一FB中。
S303、创建第二FB和第三FB。其中，第二FB可以是AmbientFB，第三FB可以是HumanFB，创建第二FB和第三FB的目的是为了在生成真实图像过程中，利用两个FB分别存储环境阴影的图像数据和运动对象的阴影的图像数据，以能够在生成真实图像过程中拦截到环境阴影的图像数据和运动对象的阴影的图像数据。
在一些示例中，步骤S303在确定开启阴影渲染后执行，即步骤S303和步骤S302可以并行执行；在一些示例中，步骤S303可以在步骤S302之前执行，此处不进行限定。
S304、将环境阴影的图像数据复制到第二FB中，实现利用第二FB对环境阴影的图像 数据的独立存储，从而在不妨碍真实图像生成的情况下，拦截到环境阴影的图像数据。在一些示例中，调用glBlitFramebuffer()函数，将环境阴影的图像数据复制到第二FB中。
S305、将应用程序绑定的FB从第一FB改为第三FB。一种可能的实现方式是，调用glBindFrameBuffer(Gl_DRAW_BUFFER,ID)函数，ID是第三FB的ID，从而将应用程序绑定的FB从第一FB改为第三FB。
在完成环境阴影的渲染后，可以渲染运动对象的阴影。若在渲染运动对象的阴影时，应用程序绑定的FB为第一FB，那么运动对象的阴影的图像数据被存储到第一FB中，运动对象的阴影的图像数据和环境阴影的图像数据混合存储到第一FB中，难以从第一FB中获取运动对象的阴影的图像数据，因此在开始渲染运动对象的阴影之前，将应用程序绑定的FB从第一FB改为第三FB，渲染运动对象的阴影时生成的运动对象的阴影的图像数据被存储到第三FB中，从而在生成真实图像过程中拦截到运动对象的阴影的图像数据。
S306、渲染运动对象的阴影，以生成运动对象的阴影的图像数据。
S307、将应用程序绑定的FB从第三FB改为第一FB，即恢复到应用程序的默认FB。如调用glBindFrameBuffer(Gl_DRAW_BUFFER,ID)函数，ID是第一FB的ID，从而将应用程序绑定的FB从第三FB改为第一FB。
S308、再次渲染运动对象的阴影，以生成运动对象的阴影的图像数据。因为应用程序绑定的FB改为第一FB，那么运动对象的阴影的图像数据被存储到第一FB中，保证应用程序的渲染结果不变，即保证应用程序默认的第一FB中存储有环境阴影的图像数据和运动对象的阴影的图像数据，在生成真实图像过程中能够从第一FB中读取到这两部分阴影的图像数据，防止真实图像缺少阴影。在一些示例中，在生成运动对象的阴影的图像数据后，可以将其复制到第一FB中，也能够保证第一FB中存储有环境阴影的图像数据和运动对象的阴影的图像数据，但是图像数据的复制效率小于渲染运动对象的阴影的效率。
在一些示例中，第一FB至第三FB可以存储图像数据中R通道的像素值，以降低对第一FB至第三FB的占用。
从上述图5所示流程可知，在生成环境阴影的图像数据后，将环境阴影的图像数据复制到第二FB中，使得第一FB和第二FB中都存储有环境阴影的图像数据。并且在生成运动对象的阴影的图像数据之前，将第三FB与应用程序绑定，这在渲染运动对象的阴影时可以将运动对象的阴影的图像数据存储到第三FB中，实现在第二FB中存储环境阴影的图像数据以及在第三FB中存储运动对象的阴影的图像数据，从而在生成真实图像过程中完成对环境阴影的图像数据和运动对象的阴影的图像数据的拦截。并且环境阴影的图像数据和运动对象的阴影的图像数据是分别存储在不同的FB中，防止这两部分图像数据重叠，造成难以区分的问题。在实现拦截环境阴影的图像数据和运动对象的阴影的图像数据的同时，能够在应用程序默认的第一FB中存储有环境阴影的图像数据和运动对象的阴影的图像数据，保证第一FB中存储的图像数据的准确度，防止真实图像缺少阴影。
应用程序的阴影渲染逻辑是：当运动对象的阴影位于环境阴影中，在真实图像中不呈现运动对象的阴影，因此通过环境阴影的图像数据和运动对象的阴影的图像数据，确定运动对象的阴影是否位于环境阴影中，以此确定出运动对象的阴影在真实图像中的位置。如图6所示，其示出了确定运动对象的阴影在真实图像中的位置的流程，可以包括以下步骤：
S401、对环境阴影的图像数据和运动对象的阴影的图像数据进行逐像素点比对计算。
在本实施例中，环境阴影的图像数据可以生成一张环境阴影纹理图像，运动对象的阴影的图像数据可以生成一张运动对象阴影纹理图像。逐像素点比对计算可以是环境阴影纹理图像和运动对象阴影纹理图像中坐标相同的像素点进行比对计算。
S402、对每个像素点，确定该像素点是否位于运动对象的阴影中，如果是，执行步骤S403，如果否，执行步骤S405。
S403、确定该像素点是否位于环境阴影中，如果否，执行步骤S404，如果是，执行步骤S405。
S404、标记该像素点处于运动对象的阴影区域中。
S405、标记该像素点没有处于运动对象的阴影区域中。
当像素点位于运动对象的阴影中但是没有位于环境阴影中，说明该像素点是运动对象的阴影中的一个像素点，则标记该像素点处于运动对象的阴影区域中。当像素点没有位于运动对象的阴影中，说明该像素点不是运动对象的阴影中的一个像素点，则标记该像素点没有处于运动对象的阴影区域中。当像素点位于运动对象的阴影中，也位于环境阴影中，说明该像素点既是运动对象的阴影中的一个像素点，也是环境阴影中的一个像素点，按照“当运动对象的阴影位于环境阴影中，在真实图像中不呈现运动对象的阴影”的原则，则标记该像素点没有处于运动对象的阴影区域中。
在一种可能的实现方式中，根据像素点的像素值，确定像素点位于运动对象的阴影还是位于环境阴影。例如hunman＝texture(humanShadow,texCoords).r，用于从运动对象阴影纹理图像humanShadow中获取坐标为texCoords的像素点在r通道(有的地方记为R通道)的像素值；ambient＝texture(ambientShadow,texCoords).r，用于从环境阴影纹理图像ambientShadow中获取坐标为texCoords的像素点在r通道的像素值。即通过hunman＝texture(humanShadow,texCoords).r和ambient＝texture(ambientShadow,texCoords).r，从运动对象阴影纹理图像humanShadow和环境阴影纹理图像ambientShadow中获取相同坐标下的像素点在r通道的像素值。
如果ambient>hunman，说明该像素点位于运动对象的阴影中，将该像素点标记为处于运动对象的阴影区域中；如果ambient≤hunman，说明该像素点位于环境阴影中，将该像素点标记为没有处于运动对象的阴影区域中。由此通过上述hunman＝texture(humanShadow,texCoords).r和ambient＝texture(ambientShadow,texCoords).r对两张纹理图像进行采样，以获取到坐标相同的每个像素点在两张纹理图像中的r通道的像素值，然后通过像素值的大小关系，确定是处于运动对象的阴影区域还是没有处于运动对象的阴影区域。
在完成所有像素点的比对后，得到一张掩膜图像，该掩膜图像表示了真实图像中运动对象的阴影区域。如在标记像素点时，将处于运动对象的阴影区域的像素点标记为1，将没有处于运动对象的阴影区域的像素点标记为0，通过1和0确定出运动对象的阴影在真实图像中的区域，从而确定出运动对象的阴影在真实图像中的位置，即确定出运动对象的阴影在真实图像的主图像中的位置。
图7示出了抠除根据运动矢量生成的运动对象的阴影的流程，以在根据运动矢量生成 第一中间图像过程中，抠除第一中间图像中运动对象的阴影为例进行说明，可以包括以下步骤：
S501、根据运动矢量对第二主图像中的像素点逐个进行图像扭曲。其中，运动矢量记录第二主图像中各像素点的移动方向和移动速度，针对第二主图像中的任一像素点，图像扭曲是指根据该像素点的移动方向和移动速度，移动该像素点。每移动一个像素点，判断该像素点是否位于运动对象的阴影中，以根据判断结果决定该像素点的像素值。
S502、根据运动对象的阴影在第二主图像中的位置，确定该像素点是否位于运动对象的阴影中，如果否，执行步骤S503，如果是，执行步骤S504。
可以理解的是：位于第二主图像的运动对象的阴影中的各像素点在移动后，仍位于运动对象的阴影中，基于该原理，对该像素点进行图像扭曲时确定该像素点是否位于第二主图像的运动对象的阴影中，如果是确定该像素点位于运动对象的阴影中，否则确定该像素点没有位于运动对象的阴影中。
在其他实现方式中，运动对象的阴影在第二主图像中的位置是可知的，根据运动矢量中与运动对象的阴影相匹配的移动方向和移动速度，能够提前确定出运动对象的阴影在第一中间图像中的位置。在完成对第二主图像中任一像素点的移动后，根据该像素点在第一中间图像中的坐标和运动对象的阴影在第一中间图像中的位置，确定该像素点是否位于第一中间图像的运动对象的阴影中。例如，该像素点在第一中间图像中的坐标与第一中间图像的运动对象的阴影中一个像素点的坐标相同，确定该像素点位于第一中间图像的运动对象的阴影中，否则确定该像素点没有位于第一中间图像的运动对象的阴影中。
又例如，根据运动对象的阴影在第一中间图像中的位置，生成掩膜图像，该掩膜图像中处于运动对象的阴影区域的像素点标记为1，将没有处于运动对象的阴影区域的像素点标记为0，在完成对第二主图像中任一像素点的移动后，根据该像素点在第一中间图像中的坐标，从掩膜图像中查找该坐标的像素点的像素值，如果位于该坐标的像素点的像素值为1，确定位于运动对象的阴影中，否则，确定没有位于运动对象的阴影中。
S503、将该像素点的像素值填充为该像素点在第二主图像中的像素值。
S504、舍弃该像素点在第二主图像中的像素值。如将该像素点的像素值设置为预设值，如预设值可以是(RGBA＝＝1.0,1.0,1.0,0.25)，以预设值来区分第一中间图像的运动对象的阴影与第一中间图像中的其他对象，从而通过预设值抠除第一中间图像中的运动对象的阴影并能够标注出第一中间图像中运动对象的阴影的位置。
S505、在完成对第二主图像中每个像素点的图像扭曲后，生成抠除运动对象的阴影的第一中间图像，从而得到一帧有环境、环境阴影、运动对象，但是抠除运动对象的阴影的第一中间图像。其中抠除运动对象的阴影可以是第一中间图像的运动对象的阴影中各像素点的像素值为预设值。
图8为本申请提供的复用真实图像中运动对象的阴影的图像数据的流程，可以包括以下步骤：
S601、依次获取抠除运动对象的阴影的第一中间图像中的各像素点。
S602、对各像素点，根据运动对象的阴影在真实图像中的位置，确定该像素点是否位于运动对象的阴影中，如果是，执行步骤S603，如果否，执行步骤S604。
此处，该像素点是否位于运动对象的阴影中是指该像素点是否位于真实图像的运动对象的阴影中，具体是指真实图像的主图像的运动对象的阴影。在一种可能的实现方式中，运动对象的阴影在真实图像中的位置可以通过位于运动对象的阴影中各像素点的坐标表示，如果抠除运动对象的阴影的第一中间图像中像素点的坐标与位于运动对象的阴影中一个像素点的坐标相同，则确定抠除运动对象的阴影的第一中间图像中的像素点位于运动对象的阴影中，否则，确定抠除运动对象的阴影的第一中间图像中的像素点没有位于运动对象的阴影中。
在一种可能的实现方式中，运动对象的阴影在真实图像中的位置通过上述图6得到的掩膜图像表示，具体通过掩膜图像中的像素值为1的像素点表示，那么在确定抠除运动对象的阴影的第一中间图像中像素点是否位于运动对象的阴影中时，利用该像素点的坐标，查找掩膜图像中位于该坐标的像素点的像素值，如果位于该坐标的像素点的像素值为1，确定位于运动对象的阴影中，否则，确定没有位于运动对象的阴影中。
S603、复用该像素点在真实图像中的像素值。即将该像素点的像素值填充为该像素点在真实图像中的像素值，因为复用像素值的像素点位于运动对象的阴影中，所以此处的复用是复用了真实图像中运动对象的阴影的图像数据。
S604、保持该像素点的像素值不变。
S605、生成复用真实图像中运动对象的阴影的图像数据的第一中间图像，从而生成具有运动对象的阴影的第一中间图像。
从上述图8可知，利用抠除运动对象的阴影的第一中间图像、运动对象的阴影在真实图像中的位置和真实图像中运动对象的阴影的图像数据，能够将真实图像中运动对象的阴影的图像数据融合到第一中间图像中，从而很简单地就融合出具有运动对象的阴影的第一中间图像，实现对真实图像中运动对象的阴影的图像数据的复用，避免继续使用环境所用的运动矢量生成运动对象的阴影，减少了预测帧存在的阴影抖动问题，降低预测帧与真实图像之间的差异，提高用户体验。
在本实施例中，具有运动对象的阴影的第一中间图像中仍存在一些像素值没有赋值的像素点(即上述待补像素点)，在生成具有运动对象的阴影的第一中间图像后，在该第一中间图像上进行图像补全处理，再进行后处理和融合，即可生成预测帧。
以游戏应用程序为例进行说明，图9中的(1)示出了用于生成预测帧的一帧真实图像，在生成该真实图像过程中，确定出人物的阴影在真实图像中的位置(如图9中的(1)的不规则图形圈住的位置)。利用运动矢量对该真实图像的环境进行图像扭曲，得到图9中的(2)所示的中间图像，该中间图像抠除了人物的阴影和人物。然后对人物进行图像扭曲，得到图9中的(3)所示的中间图像，该中间图像抠除了人物的阴影，且人物附近存在待补全区域。
对图9中的(3)所示的中间图像渲染阴影，在渲染阴影过程中，将图9中的(1)所示真实图像中人物的阴影的图像数据直接复制到图9中的(3)所示的中间图像上，得到图9中的(4)所示的图像，该图像中人物附近和人物的阴影附近都存在待补全区域。对待补全区域中的各像素点进行图像补全处理，得到图9中的(5)所示的图像。图9中的(5)所示的图像的人物的阴影是通过复用图9中的(1)所示的人物的阴影，因此两帧图像中人 物的阴影是相同的，具体是两帧图像中人物的阴影的图像数据和在图像中的位置相同。
此外，本申请还提供一种电子设备，包括：一个或多个处理器以及存储器，存储器用于存储计算机程序代码，计算机程序代码包括计算机指令，当一个或多个处理器执行计算机指令时，电子设备执行上述图像处理方法。
本申请还提供一种计算机可读存储介质，用于存储计算机程序，计算机程序被执行时，具体用于实现上述图像处理方法。
本申请还提供了一种包含指令的计算机程序产品，当该计算机程序产品在计算机或处理器上运行时，使得计算机或处理器执行上述图像处理方法。

Claims (15)

  1. 一种图像处理方法，其特征在于，应用于电子设备中，所述方法包括：
    从第一真实图像中获取第一主图像以及从第二真实图像中获取第二主图像，所述第一真实图像和所述第二真实图像是所述电子设备渲染得到的图像；
    对所述第一主图像和所述第二主图像进行运动矢量计算，得到运动矢量；
    根据所述运动矢量，对目标主图像中的环境和所述目标主图像中的运动对象进行扭曲，生成预测帧中的环境和所述预测帧中的运动对象，所述目标主图像为所述第一主图像和所述第二主图像中的一帧主图像；
    在所述预测帧中复用运动对象的阴影在所述目标主图像中的图像数据，以生成所述预测帧中的运动对象的阴影。
  2. 根据权利要求1所述的方法，其特征在于，所述在所述预测帧中复用运动对象的阴影在所述目标主图像中的图像数据包括：
    对所述预测帧中的各像素点，根据所述运动对象的阴影在所述目标主图像中的位置，确定所述像素点是否位于所述目标主图像的运动对象的阴影中；
    在确定所述像素点位于所述目标主图像的运动对象的阴影中，复用所述像素点在所述目标主图像中的像素值，所述像素点在所述目标主图像中的像素值为所述像素点在所述目标主图像中的图像数据；
    在确定所述像素点没有位于所述目标主图像的运动对象的阴影中，保持所述像素点的像素值不变。
  3. 根据权利要求1所述的方法，其特征在于，所述运动矢量用于记录所述目标主图像中各像素点的移动速度和移动方向，所述根据所述运动矢量，对目标主图像中的环境和所述目标主图像中的运动对象进行扭曲，生成预测帧中的环境和所述预测帧中的运动对象包括：
    根据目标主图像中各像素点的移动速度和移动方向，对所述目标主图像中的各像素点进行移动；
    在对所述像素点进行移动过程中，根据所述运动对象的阴影在所述目标主图像中的位置，确定所述像素点是否位于所述预测帧的运动对象的阴影中；
    在确定所述像素点位于所述预测帧的运动对象的阴影中，将所述像素点的像素值设置为预设值；
    在确定所述像素点没有位于所述预测帧的运动对象的阴影中，将所述像素点的像素值设置为所述像素点在所述目标主图像中的像素值。
  4. 根据权利要求3所述的方法，其特征在于，所述预设值包括预设透明度，所述预设透明度与所述目标主图像中其他像素点的透明度不同。
  5. 根据权利要求2至4中任意一项所述的方法，其特征在于，所述运动对象的阴影在所述目标主图像中的位置的确定过程包括：
    将环境阴影纹理图像和运动对象阴影纹理图像中位置相同的各待比对像素点进行比对，所述环境阴影纹理图像根据所述目标主图像中环境阴影的图像数据生成，所述运动对象阴影纹理图像根据所述目标主图像中运动对象的阴影的图像数据生成；
    在确定所述待比对像素点处于运动对象的阴影区域但没有处于环境阴影中，标记所述待比对像素点处于所述运动对象的阴影区域中；
    在确定所述待比对像素点处于运动对象的阴影区域且处于环境阴影中，标记所述待比对像素点没有处于所述运动对象的阴影区域中；
    根据标记为处于所述运动对象的阴影区域中的所述待比对像素点，确定所述运动对象的阴影在所述目标主图像中的位置，所述运动对象的阴影在所述目标主图像中的图像数据包括位于所述位置中的所述待比对像素点在所述目标主图像中的像素值。
  6. 根据权利要求5所述的方法，其特征在于，若所述运动对象阴影纹理图像中所述待比对像素点的像素值大于所述环境阴影纹理图像中所述待比对像素点的像素值，确定所述待比对像素点处于运动对象的阴影区域但没有处于环境阴影中；
    若所述运动对象阴影纹理图像中所述待比对像素点的像素值小于或等于所述环境阴影纹理图像中所述待比对像素点的像素值，确定所述待比对像素点处于运动对象的阴影区域且处于环境阴影中。
  7. 根据权利要求6所述的方法，其特征在于，所述待比对像素点的像素值包括所述待比对像素点的R通道值、G通道值和B通道值中的至少一个。
  8. 根据权利要求1至7中任意一项所述的方法，其特征在于，在从第一真实图像中获取第一主图像以及从第二真实图像中获取第二主图像之前，所述方法还包括：
    在渲染所述目标主图像对应的真实图像时，得到环境阴影的图像数据和所述运动对象的阴影的图像数据，所述环境阴影的图像数据和所述运动对象的阴影的图像数据用于确定所述运动对象的阴影在所述目标主图像中的位置，所述运动对象的阴影在所述目标主图像中的位置和所述运动对象的阴影的图像数据用于得到待复用的运动对象的阴影在所述目标主图像中的图像数据。
  9. 根据权利要求8所述的方法，其特征在于，所述在渲染所述目标主图像对应的真实图像时，得到环境阴影的图像数据和所述运动对象的阴影的图像数据包括：
    渲染所述真实图像的环境阴影，以生成所述环境阴影的图像数据，所述环境阴影的图像数据存储到第一帧缓冲中；
    创建第二帧缓冲和第三帧缓冲；
    将所述第一帧缓冲中存储的所述环境阴影的图像数据复制到所述第二帧缓冲中；
    将所述真实图像绑定的帧缓冲从所述第一帧缓冲改为所述第三帧缓冲；
    渲染所述真实图像的运动对象的阴影，以生成所述运动对象的阴影的图像数据，所述运动对象的阴影的图像数据存储到所述第三帧缓冲中。
  10. 根据权利要求9所述的方法，其特征在于，所述在渲染所述目标主图像对应的真实图像时，得到环境阴影的图像数据和所述运动对象的阴影的图像数据还包括：
    在渲染完所述真实图像的运动对象的阴影后，将所述真实图像绑定的帧缓冲从所述第三帧缓冲改为所述第一帧缓冲；
    再次渲染所述真实图像的运动对象的阴影，以生成所述运动对象的阴影的图像数据，所述运动对象的阴影的图像数据存储到所述第一帧缓冲中。
  11. 根据权利要求9或10所述的方法，其特征在于，所述将所述第一帧缓冲中存储的 所述环境阴影的图像数据复制到所述第二帧缓冲中包括：调用glBlitFramebuffer函数将所述第一帧缓冲中存储的所述环境阴影的图像数据复制到所述第二帧缓冲中。
  12. 根据权利要求9至11中任意一项所述的方法，其特征在于，所述将所述真实图像绑定的帧缓冲从所述第一帧缓冲改为所述第三帧缓冲包括：调用glBindFrameBuffer函数将所述真实图像绑定的帧缓冲从所述第一帧缓冲改为所述第三帧缓冲。
  13. 根据权利要求1至12中任意一项所述的方法，其特征在于，所述方法还包括：在生成所述预测帧中的运动对象的阴影后，对所述预测帧进行图像补全。
  14. 一种电子设备，其特征在于，包括：
    一个或多个处理器以及存储器；
    所述存储器用于存储计算机程序代码，所述计算机程序代码包括计算机指令，当所述一个或多个处理器执行所述计算机指令时，所述电子设备执行如权利要求1至13中任意一项所述的图像处理方法。
  15. 一种计算机可读存储介质，其特征在于，用于存储计算机程序，所述计算机程序被执行时，具体用于实现如权利要求1至13中任意一项所述的图像处理方法。
PCT/CN2023/116594 2022-11-21 2023-09-01 图像处理方法及装置 Ceased WO2024109247A1 (zh)

Priority Applications (3)

Application Number Priority Date Filing Date Title
CN202380062845.0A CN119790435A (zh) 2022-11-21 2023-09-01 图像处理方法及装置
EP23893340.2A EP4498326A4 (en) 2022-11-21 2023-09-01 IMAGE PROCESSING METHOD AND APPARATUS
US18/860,941 US20250356552A1 (en) 2022-11-21 2023-09-01 Image processing method and apparatus

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
CN202211454684.3 2022-11-21
CN202211454684.3A CN118057460A (zh) 2022-11-21 2022-11-21 图像处理方法及装置

Publications (1)

Publication Number Publication Date
WO2024109247A1 true WO2024109247A1 (zh) 2024-05-30

Family

ID=91069096

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2023/116594 Ceased WO2024109247A1 (zh) 2022-11-21 2023-09-01 图像处理方法及装置

Country Status (4)

Country Link
US (1) US20250356552A1 (zh)
EP (1) EP4498326A4 (zh)
CN (2) CN118057460A (zh)
WO (1) WO2024109247A1 (zh)

Families Citing this family (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN120166308A (zh) * 2025-03-20 2025-06-17 北京航空航天大学 一种面向真实场景的光场3d片源生成方法

Citations (7)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US9576393B1 (en) * 2014-06-18 2017-02-21 Amazon Technologies, Inc. Dynamic rendering of soft shadows for interface elements
CN109410300A (zh) * 2018-10-10 2019-03-01 苏州好玩友网络科技有限公司 一种游戏场景中的阴影处理方法及装置及终端设备
CN109672886A (zh) * 2019-01-11 2019-04-23 京东方科技集团股份有限公司 一种图像帧预测方法、装置及头显设备
US20200111195A1 (en) * 2018-10-09 2020-04-09 Valve Corporation Motion smoothing for re-projected frames
CN111028357A (zh) * 2018-10-09 2020-04-17 北京嘀嘀无限科技发展有限公司 增强现实设备的软阴影处理方法和装置
WO2022108472A1 (en) * 2020-11-20 2022-05-27 Huawei Technologies Co., Ltd Device and method for optimizing power consumption during frames rendering
CN114742934A (zh) * 2022-04-07 2022-07-12 北京字跳网络技术有限公司 图像渲染方法、装置、可读介质及电子设备

Patent Citations (7)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US9576393B1 (en) * 2014-06-18 2017-02-21 Amazon Technologies, Inc. Dynamic rendering of soft shadows for interface elements
US20200111195A1 (en) * 2018-10-09 2020-04-09 Valve Corporation Motion smoothing for re-projected frames
CN111028357A (zh) * 2018-10-09 2020-04-17 北京嘀嘀无限科技发展有限公司 增强现实设备的软阴影处理方法和装置
CN109410300A (zh) * 2018-10-10 2019-03-01 苏州好玩友网络科技有限公司 一种游戏场景中的阴影处理方法及装置及终端设备
CN109672886A (zh) * 2019-01-11 2019-04-23 京东方科技集团股份有限公司 一种图像帧预测方法、装置及头显设备
WO2022108472A1 (en) * 2020-11-20 2022-05-27 Huawei Technologies Co., Ltd Device and method for optimizing power consumption during frames rendering
CN114742934A (zh) * 2022-04-07 2022-07-12 北京字跳网络技术有限公司 图像渲染方法、装置、可读介质及电子设备

Non-Patent Citations (2)

* Cited by examiner, † Cited by third party
Title
See also references of EP4498326A4
WAN GANG, CAO XUE-FENG, LI FENG, ZHANG TING: "IBR Real-Time Shadow Technology in the Battlefield Environment Simulation", JOURNAL OF GEOMATICS SCIENCE AND TECHNOLOGY, CN, no. 01, 28 February 2007 (2007-02-28), CN, pages 18 - 21, XP009554994, ISSN: 1673-6338 *

Also Published As

Publication number Publication date
CN119790435A (zh) 2025-04-08
EP4498326A4 (en) 2025-08-20
CN118057460A (zh) 2024-05-21
EP4498326A1 (en) 2025-01-29
US20250356552A1 (en) 2025-11-20

Similar Documents

Publication Publication Date Title
WO2018072270A1 (zh) 一种图像显示增强方法及装置
CN107958480A (zh) 图像渲染方法、装置及存储介质
CN112835485A (zh) 应用界面处理方法、装置、电子设备及可读存储介质
EP4024186A1 (en) Screenshot method and terminal device
CN105278950A (zh) 用于执行视频通话增强功能的方法及其电子装置
US12153792B2 (en) Keyboard display method, foldable-screen device and computer-readable storage medium
CN114339313A (zh) 插帧方法、装置及电子设备
CN103298530B (zh) 信息处理装置
CN115129214A (zh) 一种显示设备和颜色填充方法
WO2024109247A1 (zh) 图像处理方法及装置
WO2025011490A1 (zh) 视频特效添加方法、装置、设备、存储介质和程序产品
CN102801936B (zh) 实现在屏显示的方法
WO2024017145A1 (zh) 显示方法和电子设备
US20240187725A1 (en) Photographing method and electronic device
CN114863008B (zh) 图像处理方法、装置、电子设备及存储介质
CN113822010A (zh) 内容显示方法、装置、设备及存储介质
CN107454308A (zh) 显示控制设备及其控制方法和存储介质
EP4210312A1 (en) Photographing method and electronic device
CN113453069B (zh) 一种显示设备及缩略图生成方法
CN118259811A (zh) 信息获取方法、装置、电子设备及介质
CN112995539B (zh) 一种移动终端及图像处理方法
CN115185443A (zh) 手写输入方法、装置、终端设备及计算机可读存储介质
CN115314588B (zh) 背景同步方法、装置、终端、设备、系统及存储介质
EP4603970A1 (en) Window animation processing method and electronic device
JP2015176493A (ja) 端末装置、表示プログラム及び表示方法

Legal Events

Date Code Title Description
121 Ep: the epo has been informed by wipo that ep was designated in this application

Ref document number: 23893340

Country of ref document: EP

Kind code of ref document: A1

WWE Wipo information: entry into national phase

Ref document number: 2023893340

Country of ref document: EP

ENP Entry into the national phase

Ref document number: 2023893340

Country of ref document: EP

Effective date: 20241023

WWE Wipo information: entry into national phase

Ref document number: 18860941

Country of ref document: US

WWE Wipo information: entry into national phase

Ref document number: 202380062845.0

Country of ref document: CN

WWP Wipo information: published in national office

Ref document number: 202380062845.0

Country of ref document: CN

NENP Non-entry into the national phase

Ref country code: DE

WWP Wipo information: published in national office

Ref document number: 18860941

Country of ref document: US