WO2017000765A1 - 一种目标跟踪装置及目标跟踪方法 - Google Patents
一种目标跟踪装置及目标跟踪方法 Download PDFInfo
- Publication number
- WO2017000765A1 WO2017000765A1 PCT/CN2016/085626 CN2016085626W WO2017000765A1 WO 2017000765 A1 WO2017000765 A1 WO 2017000765A1 CN 2016085626 W CN2016085626 W CN 2016085626W WO 2017000765 A1 WO2017000765 A1 WO 2017000765A1
- Authority
- WO
- WIPO (PCT)
- Prior art keywords
- motion
- target
- module
- moving target
- camera
- 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
Links
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING OR CALCULATING; COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/20—Analysis of motion
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04N—PICTORIAL COMMUNICATION, e.g. TELEVISION
- H04N23/00—Cameras or camera modules comprising electronic image sensors; Control thereof
- H04N23/60—Control of cameras or camera modules
- H04N23/62—Control of parameters via user interfaces
-
- 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
- G06T7/251—Analysis of motion using feature-based methods, e.g. the tracking of corners or segments involving models
-
- G—PHYSICS
- G06—COMPUTING OR CALCULATING; COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/20—Analysis of motion
- G06T7/254—Analysis of motion involving subtraction of images
-
- 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
-
- 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
- G06T2207/30—Subject of image; Context of image processing
- G06T2207/30232—Surveillance
Definitions
- the present invention relates to the field of electronic device technologies, and in particular, to a target tracking device and a tracking method.
- the target tracking mainly includes two aspects: first, based on a fixed static camera, analyzing the content acquired by the camera to achieve target tracking under a static background; Based on a certain type of object detection model, target detection is completed and the detection results are tracked.
- the tracking algorithm based on the fixed camera adopts the frame difference method.
- the latter is a variant method similar to the principle of the frame difference method.
- These algorithms are easy to lose the tracking target when the camera changes direction or position; and the tracking algorithm based on the object detection model needs
- the model of the object is trained in advance, and the algorithm can better detect the object that has been trained in the model, but if it is extended to multiple objects, it is necessary to train a plurality of different models in advance, in which case it will be generated.
- the present invention provides a target tracking device and a target tracking method.
- a target tracking device including a target tracking module, a motion module and a comprehensive analysis module
- the tracking algorithm is compensated by the motion parameters of the moving target in the camera, and the target tracking of multiple types of problems in a non-static background is realized. for:
- a target tracking device wherein the device is applied to tracking of a target by a camera, the device comprising:
- the target tracking module acquires an image acquired by the camera, determines the moving target according to the collected image, and finally determines an effective moving target;
- a motion module connected to the target tracking module, calculating parameter information of the current motion of the effective moving target, and determining, according to the parameter information, whether the camera needs to be rotated;
- the comprehensive analysis module is connected to the motion module and the target tracking module, and refreshes the parameter information after the camera rotates and notifies the target tracking module to continue tracking the target.
- the target tracking device wherein the target tracking module comprises:
- a moving target image acquiring module collecting a current frame in the image acquired by the camera and an image of the first two frames relative to the current frame;
- a first storage module configured to store an image acquired by the moving target image acquiring module
- the effective moving target determining module determines an effective moving target based on the image stored in the storage module.
- the target tracking device wherein the motion module specifically includes:
- a calculation module configured to calculate parameter information of the current motion of the effective moving target
- a second storage module connected to the computing module, to store the parameter information
- the determining module is connected to the storage module to determine whether the camera needs to be rotated according to the parameter information stored in the storage module.
- the parameter information includes a continuous motion time, a motion displacement, and a motion azimuth of the effective moving target.
- a target tracking method based on the above target tracking device includes: acquiring a moving image of a moving target, determining the moving target according to the moving image, and finally determining an effective moving target;
- the above-mentioned target tracking method wherein the step of acquiring a moving image of a moving target, determining the moving target according to the moving image, and finally determining the effective moving target specifically includes:
- a point that is not equal to 0 in the motion graph is represented as a motion element, and all connected motion elements are found to determine a suspected moving target;
- the above-mentioned target tracking method wherein the method for determining whether the camera needs to be rotated according to the parameter information specifically includes:
- the analysis determines that when the motion duration reaches the motion duration threshold, the motion displacement reaches the motion displacement threshold, and the motion azimuth reaches the associated motion azimuth threshold, the camera is rotated and the parameters stored in the storage module are refreshed. Information; otherwise, continue to determine the next set of parameter information for the effective moving target.
- the moving target can be accurately tracked in real time, and the moving target can be continuously tracked as soon as the camera is rotated, without re-calculating the motion detection and tracking information as in the conventional method, and being able to quickly track any moving object, A model that does not depend on objects.
- FIG. 1 is a flowchart of a moving object tracking algorithm in an embodiment of the present invention
- FIG. 2 is a schematic structural diagram of a method for tracking a moving target according to an embodiment of the present invention
- FIG. 3 is a schematic structural diagram of a method for determining an effective moving target in an embodiment of the present invention
- FIG. 4 is a schematic structural diagram of a method for determining whether a camera needs to be rotated in an embodiment of the present invention
- FIG. 5 is a schematic flow chart of an algorithm for determining whether a camera needs to be rotated in an embodiment of the present invention.
- the invention introduces a target tracking device, wherein the device is applied to the tracking of a moving target by a camera, the device mainly comprises a target tracking module, a motion module and a comprehensive analysis module, wherein:
- the target tracking module is configured to acquire an image acquired by the camera, determine the moving target according to the acquired image, and finally determine an effective moving target;
- the motion module is configured to be connected to the target tracking module, calculate parameter information of the current motion of the effective moving target, and determine, according to the parameter information, whether the camera needs to be rotated;
- the comprehensive analysis module is used to connect with the motion module and the target tracking module.
- the parameter information is refreshed and the target tracking module is notified to continue tracking the moving target.
- the target tracking module includes:
- the moving target image acquiring module collects the current frame and the image acquired by the camera with respect to the first two frames of the current frame;
- a first storage module configured to store an image acquired by the moving target image acquiring module
- An effective moving target determining module is determined according to an image stored in the storage module Effective moving target.
- the motion module specifically includes a computing module, a second storage module, and a determining module, where:
- the calculation module is configured to calculate parameter information of the current motion of the effective moving target
- the second storage module is connected to the computing module to store the parameter information
- the judging module is connected to the storage module to determine whether the camera needs to be rotated according to the parameter information stored in the storage module.
- the parameter information includes continuous motion time, motion displacement and motion azimuth of the effective moving target.
- the image is first acquired, the moving target is determined, the effective moving target is continuously determined, and then the parameter information of the moving target is calculated.
- the parameter information reaches the set threshold, the camera is rotated and the parameter information is updated and stored in The location information of the camera in the memory, the specific method is:
- the method specifically includes:
- the step of acquiring a moving image of a moving target, determining the moving target according to the moving image, and finally determining the effective moving target specifically includes:
- a point whose image is not equal to 0 is represented as a motion element, and all connected motion elements are found to determine a suspected moving target;
- the second frame minus the third frame obtains a difference image img2, and img1 and img2 are superimposed to obtain a moving image; then the point in the moving image that is not equal to 0 represents a motion element, and all connected motion elements are found, and these interconnections are connected.
- the motion element constitutes a suspected moving target; finally, the motion intensity of all suspected moving targets is calculated.
- the exercise intensity the number of motion elements / the rectangular area of the suspected target, the greater the exercise intensity value, the richer the motion information of the region. Conversely, the smaller the exercise intensity, the more sparse the motion information of the region is, and a motion intensity threshold is defined in advance. When the motion intensity of the suspected moving target is less than the threshold, we filter it and discard the exercise intensity less than the exercise intensity. Set the threshold of the motion element to get the effective motion target.
- the method for determining whether the camera needs to be rotated according to the parameter information specifically includes:
- the analysis determines that when the motion duration reaches the motion duration threshold, the motion displacement reaches the motion displacement threshold, and the motion azimuth reaches the associated motion azimuth threshold, the camera is rotated and the parameter information stored in the storage module is refreshed; otherwise, continue Determine the next set of parameter information for the effective moving target.
- the continuous motion time of the moving target reaches a set continuous motion time threshold. If not, the tracking of the moving target is ended, and the next motion is started.
- Target if the set motion duration threshold is reached, it is determined whether the motion displacement of the moving target reaches a set motion displacement threshold, and if not, the tracking of the moving target is ended, and the tracking of the next moving target is started, if
- the set motion displacement threshold determines whether the moving azimuth of the moving target reaches the set moving azimuth threshold. If the set moving azimuth threshold is not reached, the tracking of the moving target is ended, and the next moving target is started. Tracking, if the set azimuth threshold is reached, rotate the camera.
- the camera After the camera completes one rotation, it needs to update the parameter information of the tracking target stored in the storage module, and the dynamic information of the camera is also updated in time after one rotation, such that it can prevent the next camera rotation from being deviated, and the specific performance If: the next target moves forward toward the camera's steering direction, it may cause the next camera to turn too fast and the angle is too large; if the next target moves in the opposite direction of the camera, the latest tracking window may rotate with the camera.
- the front tracking windows coincide, or cause the motion displacement to be too small without steering, tracking lag or loss.
- the present invention constructs a moving target tracking device including a target tracking module, a motion module and a comprehensive analysis module, obtains a moving target by acquiring an image acquired by the camera, and finally determines an effective moving target, and then calculates the effective moving target.
- the parameter information of the motion it is judged whether the camera needs to be rotated; and the parameter information stored in the memory is updated after the camera is rotated.
- the moving target can be accurately and accurately tracked in real time and after the camera completes the rotation.
- the ability to continue to track moving targets eliminates the need to re-execute motion detection and tracking information accumulation as in the conventional method, and the present technical solution is capable of quickly tracking any moving object without relying on any object model.
Landscapes
- Engineering & Computer Science (AREA)
- Multimedia (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Human Computer Interaction (AREA)
- Signal Processing (AREA)
- Image Analysis (AREA)
- Studio Devices (AREA)
- Closed-Circuit Television Systems (AREA)
Abstract
本发明涉及电子设备技术领域,尤其涉及一种目标跟踪装置及跟踪方法,通过构建一包括目标跟踪模块、运动模块和综合分析模块的运动目标跟踪装置,通过从摄像头中获取运动目标的图像确定运动目标并最终确定有效运动目标,然后计算该有效运动目标本次运动的参数信息并根据该参数信息判断是否需要转动摄像头;并在摄像头转动完成后更新存储在存储器中的参数信息,通过本技术方案,能精准实时有效跟踪运动目标并在摄像头完成转动后能够继续跟踪运动目标,无需像传统方法一样重新进行一次运动检测与跟踪信息积累,而且本技术方案能够快速跟踪任何运动的物体,不依赖于任何物体模型。
Description
本发明涉及电子设备技术领域,尤其涉及一种目标跟踪装置及跟踪方法。
目前在图像处理中已经实现了目标跟踪的功能,该目标跟踪主要包括两个方面:第一、基于固定的静态摄像头,对摄像头获取的内容进行分析,实现静态背景下的目标跟踪;第二、基于某一类物体检测模型,完成目标检测,并对检测结果进行跟踪。
基于固定摄像头的跟踪算法采用的是帧差法后者是与帧差法原理相似的变种方法,这些算法在摄像头改变方向或者位置的情况下容易丢失跟踪目标;而基于物体检测模型的跟踪算法需要事先训练好物体的模型,该算法对于该类已经训练好模型的物体能够较好的检测效果,但是如果扩展到多种物体,就需要事先训练多个不同的模型,在这种情况下会产生较大的工作量,并且对目标跟踪装置的计算性能有较高的要求。因此,如何更好的解决多类物体在非静态摄像头的动态背景下的目标跟踪问题成为本领域技术人员面临的一大难题。
发明内容
针对上述问题,本发明提出一种目标跟踪装置及目标跟踪方法,
通过构建一包括目标跟踪模块、运动模块和综合分析模块的目标跟踪装置,通过摄像头中运动目标的运动参数对跟踪算法进行补差,实现多类问题在非静态背景下的目标跟踪,该技术方案具体为:
一种目标跟踪装置,其中,所述装置应用于摄像头对目标的跟踪中,所述装置包括:
目标跟踪模块,获取摄像头采集到的图像,根据采集到的图像确定所述运动目标并最终确定有效运动目标;
运动模块,与所述目标跟踪模块连接,计算所述有效运动目标本次运动的参数信息,根据所述参数信息判断是否需要转动摄像头;
综合分析模块,与所述运动模块和所述目标跟踪模块连接,当所述摄像头转动后刷新所述参数信息并通知所述目标跟踪模块继续跟踪目标。
上述目标跟踪装置,其中,所述目标跟踪模块包括:
运动目标图像获取模块,
运动目标图像获取模块,采集所述摄像头所获取到的图像中当前帧以及相对于当前帧的前两帧图像;
第一存储模块,存储所述运动目标图像获取模块获取的图像;
有效运动目标确定模块,根据存储于所述存储模块中的图像确定有效运动目标。
上述目标跟踪装置,其中,所述运动模块具体包括:
计算模块,计算所述有效运动目标本次运动的参数信息;
第二存储模块,与所述计算模块连接,以储存所述参数信息;
判断模块,与所述存储模块连接,以根据存储模块中储存的所述参数信息判断是否需要转动摄像头。
上述目标跟踪装置,其中,所述参数信息包括所述有效运动目标的持续运动时间、运动位移和运动方位角。
一种基于上述目标跟踪装置的目标跟踪方法,其中,所述方法包括:获取运动目标的运动图像,根据所述运动图像确定所述运动目标并最终确定有效运动目标;
计算所述有效运动目标本次运动的参数信息,根据所述参数信息判断是否需要转动摄像头;
若不需要转动摄像头,则继续等待下一条参数信息;若需要转动摄像头,则转动摄像头并刷新所述存储模块中储存的参数信息。
上述目标跟踪方法,其中,所述获取运动目标的运动图像,根据所述运动图像确定所述运动目标并最终确定有效运动目标的步骤具体包括:
获取第一帧、第二帧和第三帧从摄像头中采集的图像信息,得到所述运动目标的运动图;
在运动图中不等于0的点表示为一个运动元,找到所有相互连接的运动元,确定疑似运动目标;
计算所述疑似运动目标的运动强度,舍弃运动强度小于一阈值的运动元,确定有效运动目标。
上述目标跟踪方法,其中,所述根据所述参数信息判断是否需要转动摄像头的方法具体包括:
分别设定有效运动目标的持续运动时间阈值、运动位移阈值和运动方位角阈值并将其存储于一存储模块中;
计算所述有效运动目标的有效运动目标的持续运动时间、运动位移和运动方位角并将其放置于所述存储模块中;
分析判断,当所述运动持续时间达到所述运动持续时间阈值、所述运动位移达到所述运动位移阈值且所述运动方位角达到所属运动方位角阈值时转动摄像头并刷新存储模块中储存的参数信息;否则,继续判断有效运动目标的下一组参数信息。
本发明具有的优点以及能达到的有益效果:
通过采用本发明的技术方案,精准实时跟踪运动目标,且在摄像头转动后能够尽快继续跟踪运动目标,无须像传统方法一样重新进行运动检测与跟踪信息的积累,并且能够快速跟踪任何运动的物体,不依赖物体的模型。
通过阅读参照以下附图对非限制性实施例所作的详细描述,本发明及其特征外形和优点将会变得更加明显。在全部附图中相同的标记指示相同的部分。并未可以按照比例绘制附图,重点在于示出本发明的主旨。
图1是本发明实施例中运动目标跟踪算法流程图;
图2是本发明实施例跟踪运动目标的方法结构示意图;
图3是本发明实施例中确定有效运动目标的方法结构示意图;
图4是本发明实施例中判断是否需要转动摄像头的方法结构示意图;
图5是本发明实施例中判断是否需要转动摄像头的算法流程示意图。
下面结合附图和具体的实施例对本发明作进一步的说明,但是不作为本发明的限定。
本发明介绍一种目标跟踪装置,其中,该装置应用于摄像头对运动目标的跟踪中,该装置主要包括目标跟踪模块、运动模块和综合分析模块,其中:
目标跟踪模块用于获取摄像头获取的图像,并根据获取到的图像确定该运动目标并最终确定有效运动目标;
运动模块用来与该目标跟踪模块连接,计算该有效运动目标本次运动的参数信息,根据该参数信息判断是否需要转动摄像头;
综合分析模块用来与运动模块和目标跟踪模块连接,当摄像头转动后刷新参数信息并通知目标跟踪模块继续跟踪运动目标。
作为本发明一个优选实施例,目标跟踪模块包括:
运动目标图像获取模块,采集当前帧以及相对于当前帧的前两帧摄像头获取到的图像;
第一存储模块,存储所述运动目标图像获取模块获取的图像;
有效运动目标确定模块,根据存储于所述存储模块中的图像确定
有效运动目标。
作为本发明一个优选实施例,运动模块具体包括计算模块、第二存储模块和判断模块,其中:
计算模块用于计算有效运动目标本次运动的参数信息;
第二存储模块与该计算模块连接,以储存该参数信息;
判断模块与该存储模块连接,以根据存储模块中储存的该参数信息判断是否需要转动摄像头。
在此基础上,进一步的,参数信息包括有效运动目标的持续运动时间、运动位移和运动方位角。
参见图1所示结构,首先获取图像,确定运动目标,继续确定有效运动目标,然后计算运动目标的参数信息,当各项参数信息达到设定的阈值时便转动摄像头并更新参数信息以及存储在存储器中的摄像头的位置信息,具体方法为:
参见图2所示结构示意图,其中,该方法具体包括:
获取摄像头采集到的图像,根据获取的图像确定运动目标并最终确定有效运动目标;
计算该有效运动目标本次运动的参数信息,根据该参数信息判断是否需要转动摄像头;
若不需要转动摄像头,则继续等待下一条参数信息;若需要转动摄像头,则转动摄像头并刷新该存储模块中储存的参数信息,而且存储器中储存有摄像头的位置信息,摄像头每转动一次存储于存储器中的摄像头的位置信息更新一次,下一次转动时便以上一次转动后的位
置信息为参照。
参见图3所示结构示意图,在本发明一个优选实施例中,其中,获取运动目标的运动图像,根据该运动图像确定该运动目标并最终确定有效运动目标的步骤具体包括:
获取第一帧、第二帧和第三帧从摄像头中采集的图像信息,得到该运动目标的运动图像(定义当前帧为第三帧,则前两帧根据时间先后分别定义为第一帧和第二帧);
在运动图中将图像不等于0的点表示为一个运动元,找到所有相互连接的运动元,确定疑似运动目标;
计算该疑似运动目标的运动强度,舍弃运动强度小于一阈值的运动元,确定有效运动目标。
作为本发明一个具体实施例,首先获取当前帧(第三帧)、第一帧和第二帧从摄像头中采集到的图像信息,用第一帧减第三帧获得一个差值图像img1,用第二帧减第三帧获得一个差值图像img2,将img1和img2叠加得到一个运动图像;然后将运动图像中不等于0的点表示一个运动元,找到所有相互连接的运动元,这些相互连接的运动元组成一个疑似运动目标;最后,计算所有疑似运动目标的运动强度,按照公式运动强度=运动元数量/疑似目标的矩形面积,运动强度值越大,表示这个区域的运动信息越丰富,反之,运动强度越小,表示该区域的运动信息越稀疏,且事先定义一个运动强度阈值,当疑似运动目标的运动强度小于该阈值时,我们将其过滤到,及舍弃运动强度小于运动强度预设阈值的运动元,得到有效运动目标。
参见图4所示结构示意图,作为本发明一个优选实施例,其中,根据该参数信息判断是否需要转动摄像头的方法具体包括:
分别设定有效运动目标的持续运动时间阈值、运动位移阈值和运动方位角阈值并将其存储于第二存储模块中;
计算该有效运动目标的有效运动目标的持续运动时间、运动位移和运动方位角并将其放置于第二存储模块中;
分析判断,当该运动持续时间达到该运动持续时间阈值、该运动位移达到该运动位移阈值且该运动方位角达到所属运动方位角阈值时转动摄像头并刷新存储模块中储存的参数信息;否则,继续判断有效运动目标的下一组参数信息。
参见图5所示流程图,跟踪一个运动目标时,首先判断该运动目标的持续运动时间是否达到设定的持续运动时间阈值,如果没有达到,则结束对本运动目标的跟踪,开始跟踪下一个运动目标;如果达到设定的运动持续时间阈值,判断该运动目标的运动位移是否达到设定的运动位移阈值,如果没有达到,则结束本运动目标的跟踪,开始下一个运动目标的跟踪,如果达到设定的运动位移阈值,则判断运动目标的运动方位脚是否达到设定的运动方位角阈值,如果未达到设定的运动方位角阈值,则结束本运动目标的跟踪,开始下一个运动目标的跟踪,如果达到设定的运动方位角阈值,则转动摄像头。
其中,如果有一个极速的物体从摄像头前方飞过,虽然移动的距离够长,但是没有实际的跟踪必要,而设定持续时间阈值能有效防止在这种情况下摄像头转动;一个物体只是在摄像头前方轻微抖动,虽
然持续时间够长,但是运动位移达不到运动位移阈值,即设定运动位移阈值能有效避免此情况下的摄像头转动;被跟踪的目标在摄像头前方进行微小徘徊运动时,根据运动方位角阈值判断不需要转动摄像头。
在摄像头完成一次转动后,需要更新存储于存储模块中的跟踪目标的参数信息,并且,摄像头的动态信息在一次转动后也是及时更新的,这样的能放防止下一次摄像头转动有偏差,具体表现为:如果下一次目标朝摄像头的转向方向继续前进,可能会导致下一次摄像头转向速度过大,角度过大;如果下一次目标朝摄像头的反方向运动,但是最新的跟踪窗口可能会与摄像头转动前的跟踪窗口重合,或者导致运动位移过小而不进行转向,跟踪滞后或者丢失。
综上所述,本发明通过构建一包括目标跟踪模块、运动模块和综合分析模块的运动目标跟踪装置,通过获取摄像头采集到的图像确定运动目标并最终确定有效运动目标,然后计算该有效运动目标本次运动的参数信息并根据该参数信息判断是否需要转动摄像头;并在摄像头转动完成后更新存储在存储器中的参数信息,通过本技术方案,能精准实时有效跟踪运动目标并在摄像头完成转动后能够继续跟踪运动目标,无需像传统方法一样重新进行一次运动检测与跟踪信息积累,而且本技术方案能够快速跟踪任何运动的物体,不依赖于任何物体模型。
本领域技术人员应该理解,本领域技术人员在结合现有技术以及上述实施例可以实现所述变化例,在此不做赘述。这样的变化例并不
影响本发明的实质内容,在此不予赘述。
以上对本发明的较佳实施例进行了描述。需要理解的是,本发明并不局限于上述特定实施方式,其中未尽详细描述的设备和结构应该理解为用本领域中的普通方式予以实施;任何熟悉本领域的技术人员,在不脱离本发明技术方案范围情况下,都可利用上述揭示的方法和技术内容对本发明技术方案作出许多可能的变动和修饰,或修改为等同变化的等效实施例,这并不影响本发明的实质内容。因此,凡是未脱离本发明技术方案的内容,依据本发明的技术实质对以上实施例所做的任何简单修改、等同变化及修饰,均仍属于本发明技术方案保护的范围。
Claims (7)
- 一种目标跟踪装置,其特征在于,所述装置应用于摄像头对目标的跟踪,所述装置包括:目标跟踪模块,获取摄像头采集到的图像,根据采集到的图像确定所述运动目标并最终确定有效运动目标;运动模块,与所述目标跟踪模块连接,计算所述有效运动目标本次运动的参数信息,根据所述参数信息判断是否需要转动摄像头;综合分析模块,与所述运动模块和所述目标跟踪模块连接,当所述摄像头转动后刷新所述参数信息并通知所述目标跟踪模块继续跟踪目标。
- 如权利要求1所述目标跟踪装置,其特征在于,所述目标跟踪模块包括:运动目标图像获取模块,采集所述摄像头所获取到的图像中当前帧以及相对于当前帧的前两帧图像;第一存储模块,存储所述运动目标图像获取模块获取的图像;有效运动目标确定模块,根据存储于所述存储模块中的图像确定有效运动目标。
- 如权利要求1所述目标跟踪装置,其特征在于,所述运动模块具体包括:计算模块,计算所述有效运动目标本次运动的参数信息;第二存储模块,与所述计算模块连接,以储存所述参数信息;判断模块,与所述存储模块连接,以根据存储模块中储存的所述参数信息判断是否需要转动摄像头。
- 如权利要求3所述目标跟踪装置,其特征在于,所述参数信息包括所述有效运动目标的持续运动时间、运动位移和运动方位角。
- 一种目标跟踪方法,其特征在于,所述方法基于权利要求1-4中任意一项权利要求,应用于运动目标的跟踪中,所述方法包括:获取运动目标的运动图像,根据所述运动图像确定所述运动目标并最终确定有效运动目标;计算所述有效运动目标本次运动的参数信息,根据所述参数信息判断是否需要转动摄像头;若不需要转动摄像头,则继续等待下一条参数信息;若需要转动摄像头,则转动摄像头并刷新所述存储模块中储存的参数信息。
- 如权利要求5所述目标跟踪方法,其特征在于,所述获取运动目标的运动图像,根据所述运动图像确定所述运动目标并最终确定有效运动目标的步骤具体包括:获取第一帧、第二帧和第三帧从摄像头中采集的图像信息,得到所述运动目标的运动图;运动图中不等于0的点表示为一个运动元,找到所有相互连接的运动元,确定疑似运动目标;计算所述疑似运动目标的运动强度,舍弃运动强度小于一阈值的运动元,确定有效运动目标。
- 如权利要求5所述目标跟踪方法,其特征在于,所述根据所述参数信息判断是否需要转动摄像头的方法具体包括:分别设定有效运动目标的持续运动时间阈值、运动位移阈值和运 动方位角阈值并将其存储于一存储模块中;计算所述有效运动目标的有效运动目标的持续运动时间、运动位移和运动方位角并将其放置于所述存储模块中;分析判断,当所述运动持续时间达到所述运动持续时间阈值、所述运动位移达到所述运动位移阈值且所述运动方位角达到所属运动方位角阈值时转动摄像头并刷新存储模块中储存的参数信息;否则,继续判断有效运动目标的下一组参数信息。
Priority Applications (3)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| EP16817129.6A EP3319042B1 (en) | 2015-06-30 | 2016-06-13 | Target tracking device and target tracking method |
| JP2017567762A JP6657270B2 (ja) | 2015-06-30 | 2016-06-13 | 目標追跡装置及び目標追跡方法 |
| US15/739,297 US10506150B2 (en) | 2015-06-30 | 2016-06-13 | Target tracking device and target tracking method |
Applications Claiming Priority (2)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201510387049.1A CN106327517B (zh) | 2015-06-30 | 2015-06-30 | 一种目标跟踪装置及目标跟踪方法 |
| CN201510387049.1 | 2015-06-30 |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| WO2017000765A1 true WO2017000765A1 (zh) | 2017-01-05 |
Family
ID=57607839
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| PCT/CN2016/085626 Ceased WO2017000765A1 (zh) | 2015-06-30 | 2016-06-13 | 一种目标跟踪装置及目标跟踪方法 |
Country Status (6)
| Country | Link |
|---|---|
| US (1) | US10506150B2 (zh) |
| EP (1) | EP3319042B1 (zh) |
| JP (1) | JP6657270B2 (zh) |
| CN (1) | CN106327517B (zh) |
| TW (1) | TWI587113B (zh) |
| WO (1) | WO2017000765A1 (zh) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111640141A (zh) * | 2020-05-20 | 2020-09-08 | 山东神戎电子股份有限公司 | 一种基于热成像的低慢目标定位跟踪方法 |
Families Citing this family (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US10430641B2 (en) * | 2017-03-08 | 2019-10-01 | GM Global Technology Operations LLC | Methods and systems for object tracking using bounding boxes |
| CN109559330B (zh) * | 2017-09-25 | 2021-09-10 | 北京金山云网络技术有限公司 | 运动目标的视觉跟踪方法、装置、电子设备及存储介质 |
| CN107992677B (zh) * | 2017-11-29 | 2020-11-17 | 中国人民解放军火箭军工程大学 | 基于惯导信息和亮度修正的红外弱小移动目标跟踪方法 |
| CN109685062B (zh) * | 2019-01-02 | 2023-07-25 | 南方科技大学 | 一种目标检测方法、装置、设备及存储介质 |
| CN109859200B (zh) * | 2019-02-15 | 2022-03-18 | 哈尔滨工程大学 | 一种基于背景分析的低空慢速无人机快速检测方法 |
| CN111145217A (zh) * | 2019-12-27 | 2020-05-12 | 湖南华诺星空电子技术有限公司 | 一种基于kcf的无人机跟踪方法 |
| CN115442521B (zh) * | 2022-08-08 | 2023-10-27 | 杭州华橙软件技术有限公司 | 摄像控制方法及相关装置、摄像机和存储介质 |
| CN116744102B (zh) * | 2023-06-19 | 2024-03-12 | 北京拙河科技有限公司 | 一种基于反馈调节的球机跟踪方法及装置 |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN1402551A (zh) * | 2001-08-07 | 2003-03-12 | 三星电子株式会社 | 自动跟踪移动目标的设备和方法 |
| CN1937766A (zh) * | 2005-09-20 | 2007-03-28 | 富士能株式会社 | 监视摄像机装置及监视摄像机系统 |
| CN101572804A (zh) * | 2009-03-30 | 2009-11-04 | 浙江大学 | 多摄像机智能控制方法及装置 |
| CN102307297A (zh) * | 2011-09-14 | 2012-01-04 | 镇江江大科茂信息系统有限责任公司 | 一种多方位视频目标跟踪检测的智能监控系统 |
| CN103237195A (zh) * | 2013-04-07 | 2013-08-07 | 南京理工大学 | 目标图像实时传输与跟踪系统 |
Family Cites Families (18)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN1141427A (zh) * | 1995-12-29 | 1997-01-29 | 西安交通大学 | 基于图像识别的运动物体测量方法 |
| JP4819380B2 (ja) * | 2004-03-23 | 2011-11-24 | キヤノン株式会社 | 監視システム、撮像設定装置、制御方法、及びプログラム |
| KR100837435B1 (ko) * | 2006-10-25 | 2008-06-12 | 삼성전자주식회사 | 촬영장치 및 이의 대상 추적제어방법 |
| US8189905B2 (en) * | 2007-07-11 | 2012-05-29 | Behavioral Recognition Systems, Inc. | Cognitive model for a machine-learning engine in a video analysis system |
| CN101378454B (zh) * | 2007-08-31 | 2010-06-16 | 鸿富锦精密工业(深圳)有限公司 | 摄像设备及其拍摄方法 |
| JP4982407B2 (ja) * | 2008-02-22 | 2012-07-25 | 株式会社東芝 | 移動体画像追尾装置及び方法 |
| JP2010154391A (ja) * | 2008-12-26 | 2010-07-08 | Panasonic Corp | 自動追尾カメラ装置 |
| US8116527B2 (en) * | 2009-10-07 | 2012-02-14 | The United States Of America As Represented By The Secretary Of The Army | Using video-based imagery for automated detection, tracking, and counting of moving objects, in particular those objects having image characteristics similar to background |
| JP2012133759A (ja) * | 2010-11-29 | 2012-07-12 | Canon Inc | 侵入物体の検知を行うことができる物体追尾装置、物体追尾方法及び記憶媒体 |
| JP5818487B2 (ja) * | 2011-04-11 | 2015-11-18 | キヤノン株式会社 | カメラ制御装置、カメラ制御方法 |
| TWM426046U (en) * | 2011-05-10 | 2012-04-01 | Chiuan Yan Technology Co Ltd | Active target tracking microphotography module |
| CN103974028A (zh) * | 2013-01-30 | 2014-08-06 | 由田新技股份有限公司 | 人员激烈行为侦测方法 |
| JP6182332B2 (ja) * | 2013-03-08 | 2017-08-16 | 株式会社デンソーウェーブ | 監視装置の制御方法 |
| TWI509568B (zh) * | 2013-10-25 | 2015-11-21 | Univ Nat Kaohsiung Applied Sci | 偵測多移動目標之方法 |
| CN203951553U (zh) * | 2014-05-05 | 2014-11-19 | 深圳市莫孚康技术有限公司 | 一种摄影机自动跟焦装置 |
| US9354794B2 (en) * | 2014-07-07 | 2016-05-31 | Google Inc. | Method and system for performing client-side zooming of a remote video feed |
| KR102558473B1 (ko) * | 2016-09-30 | 2023-07-25 | 삼성전자주식회사 | 이미지를 표시하기 위한 방법 및 그 전자 장치 |
| US10607098B2 (en) * | 2017-11-13 | 2020-03-31 | Aupera Technologies, Inc. | System of a video frame detector for video content identification and method thereof |
-
2015
- 2015-06-30 CN CN201510387049.1A patent/CN106327517B/zh active Active
-
2016
- 2016-06-13 EP EP16817129.6A patent/EP3319042B1/en active Active
- 2016-06-13 US US15/739,297 patent/US10506150B2/en active Active
- 2016-06-13 JP JP2017567762A patent/JP6657270B2/ja active Active
- 2016-06-13 WO PCT/CN2016/085626 patent/WO2017000765A1/zh not_active Ceased
- 2016-06-29 TW TW105120415A patent/TWI587113B/zh not_active IP Right Cessation
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN1402551A (zh) * | 2001-08-07 | 2003-03-12 | 三星电子株式会社 | 自动跟踪移动目标的设备和方法 |
| CN1937766A (zh) * | 2005-09-20 | 2007-03-28 | 富士能株式会社 | 监视摄像机装置及监视摄像机系统 |
| CN101572804A (zh) * | 2009-03-30 | 2009-11-04 | 浙江大学 | 多摄像机智能控制方法及装置 |
| CN102307297A (zh) * | 2011-09-14 | 2012-01-04 | 镇江江大科茂信息系统有限责任公司 | 一种多方位视频目标跟踪检测的智能监控系统 |
| CN103237195A (zh) * | 2013-04-07 | 2013-08-07 | 南京理工大学 | 目标图像实时传输与跟踪系统 |
Non-Patent Citations (1)
| Title |
|---|
| See also references of EP3319042A4 * |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111640141A (zh) * | 2020-05-20 | 2020-09-08 | 山东神戎电子股份有限公司 | 一种基于热成像的低慢目标定位跟踪方法 |
| CN111640141B (zh) * | 2020-05-20 | 2023-10-24 | 山东神戎电子股份有限公司 | 一种基于热成像的低慢目标定位跟踪方法 |
Also Published As
| Publication number | Publication date |
|---|---|
| HK1231613A1 (zh) | 2017-12-22 |
| JP2018523394A (ja) | 2018-08-16 |
| JP6657270B2 (ja) | 2020-03-04 |
| CN106327517A (zh) | 2017-01-11 |
| US10506150B2 (en) | 2019-12-10 |
| EP3319042B1 (en) | 2020-10-07 |
| CN106327517B (zh) | 2019-05-28 |
| US20180191950A1 (en) | 2018-07-05 |
| TWI587113B (zh) | 2017-06-11 |
| EP3319042A1 (en) | 2018-05-09 |
| TW201701094A (zh) | 2017-01-01 |
| EP3319042A4 (en) | 2019-01-02 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| WO2017000765A1 (zh) | 一种目标跟踪装置及目标跟踪方法 | |
| US10769480B2 (en) | Object detection method and system | |
| Akolkar et al. | Real-time high speed motion prediction using fast aperture-robust event-driven visual flow | |
| CN108010067B (zh) | 一种基于组合判断策略的视觉目标跟踪方法 | |
| CN110692083B (zh) | 用于动态视觉传感器的块匹配光流和立体视觉 | |
| CN110827321B (zh) | 一种基于三维信息的多相机协同的主动目标跟踪方法 | |
| CN110390685B (zh) | 一种基于事件相机的特征点跟踪方法 | |
| CN112802096A (zh) | 实时定位和建图的实现装置和方法 | |
| WO2018068771A1 (zh) | 目标跟踪方法、系统、电子设备和计算机存储介质 | |
| TWI615026B (zh) | 基於人體資訊的機器人監控系統及方法 | |
| JP2017526082A (ja) | 動作推定方法、移動体、およびプロセッサに動作推定方法を実行させるコンピュータプログラムコードでコード化された非一時的コンピュータ可読媒体 | |
| CN105678808A (zh) | 运动目标跟踪方法及装置 | |
| Milford et al. | Towards visual slam with event-based cameras | |
| KR20210101286A (ko) | 장면 내 물체를 추적하는 방법 | |
| CN111627049B (zh) | 高空抛物的确定方法、装置、存储介质及处理器 | |
| CN104331901A (zh) | 一种基于tld的多视角目标跟踪装置及方法 | |
| CN105975923B (zh) | 用于跟踪人体对象的方法和系统 | |
| CN108133487A (zh) | 视频跨区域单人体姿态目标检测提取方法 | |
| Bisulco et al. | Fast motion understanding with spatiotemporal neural networks and dynamic vision sensors | |
| JP7484924B2 (ja) | 撮像装置、画像処理システム、画像処理方法及びプログラム | |
| CN112393721B (zh) | 相机位姿估计方法 | |
| CN102881025A (zh) | 一种多运动目标的检测方法 | |
| HK1231613B (zh) | 一种目标跟踪装置及目标跟踪方法 | |
| Yamane et al. | MVTrajecter: Multi-View Pedestrian Tracking with Trajectory Motion Cost and Trajectory Appearance Cost | |
| WO2015129152A1 (ja) | 画像認識システムおよび半導体集積回路 |
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: 16817129 Country of ref document: EP Kind code of ref document: A1 |
|
| ENP | Entry into the national phase |
Ref document number: 2017567762 Country of ref document: JP Kind code of ref document: A |
|
| NENP | Non-entry into the national phase |
Ref country code: DE |