JPH0332953B2 - - Google Patents

Description

[Detailed Description of the Invention] [Technical Field of the Invention] The present invention provides a grayscale image to be processed, recognized, measured, etc. by a computer by using a still image multiple times from an image input device such as an ITV camera. The present invention relates to a method for inputting and storing images obtained by inputting and storing images.

[Technical background of the invention and its problems]

Random noise generated from illumination light sources, photoelectric converters, amplifiers, etc. is superimposed on images input from ITV cameras and the like. A method to reduce this random noise is to increase the gradation by inputting and adding the same image multiple times, that is, increase the number of gradations, and then divide by the number of inputs to calculate the average value of the corresponding pixels. There are known methods to find. This is based on the principle that the added value SN ratio of an observation target that includes a certain amount of noise increases by a factor of √ as the number of observations n increases. (References, Hiroshi Ozaki, Keiji Taniguchi, Hideo Ogawa: Image processing - from its basics to applications, Kyoritsu Shuppan, 1984) High gradation is also necessary to recognize and measure subtle density changes.

In order to achieve this high gradation, when the input gradation from the ITV camera is normally 8 bits, for example, in order to input and add twice, an extra 1 bit,
Two extra bits are required to input and add three or four times. Generally, in order to input and add data n times, an extra number of bits is required, which is an integer N obtained by rounding up (log n)/(log 2), so a capacity of 8+N bits is ultimately required. However, usually ITV
Since the gradation of the camera and the gradation of the image memory are the same, it was not possible to increase the gradation by the above-mentioned addition. In particular, there is a method of using a large-capacity image memory for only one side in order to realize the addition. However, in this case, only a part of the memory is special, which makes it uneconomical to design and manufacture the image processing device.

[Purpose of the invention]

An object of the present invention is to solve the uneconomical design and manufacturing problems caused by using a large-capacity image memory on one side to input and store addition values, and to ensure that the gradation of each image memory matches the gradation of an ITV camera. The object of the present invention is to provide a method for inputting and storing a high-gradation addition image in a commonly used image processing apparatus that is generally equally provided, thereby making it economical.

[Summary of the invention]

In the present invention, when the image is still, the initial image is input and stored in the first image memory, and the initial image and the second image are stored.
The main feature is that the added value of the difference between the images after the first image is input and stored in the third image memory, and this type of method has not been found in the prior art.

[Embodiments of the invention]

First, FIG. 1 shows the system configuration of an image processing apparatus that has been widely used in the past and to which the method of the present invention is applied. A system processor 2 that controls the entire system to which a console CRT 1 is connected, a main memory 3, and an ITV camera 4 as an example of an image input device are connected and
An image processing processor 7 including an inter-image arithmetic unit 5 capable of subtraction, multiplication, and division and a grayscale image memory 6 capable of inputting and storing gradation information of at least three planes is connected via a data bus 8. Briefly explaining the operation of this apparatus, the system processor 2 controls the image processing processor 7 according to a program input from the console CRT1. The image processing processor 7 inputs images from the ITV camera 4 and stores them in the image memory 6, or the image processing unit 5
The weighted addition between the images in the image memory 6 is calculated by
Perform subtraction, multiplication, and division. Using such a device,
We invented the following image input/storage method.

FIG. 2 is a flowchart showing an embodiment of the present invention. First, for a still image, input initial image ₁ (x, y) into the first image memory in step 9.
Remember. Next, in step 10, the second and subsequent images (the i-th image _i (x, y)) are input into the second image memory and are sequentially updated and stored. Next, step
At step 11, the value in the second image memory is subtracted from the value in the first image memory to obtain ( ₁ (x, y) - _i (x, y)), which is added and stored in the third image memory. These steps 10 and 11 are combined with the condition i=n of step 12.
This is repeated until the number of image inputs i reaches a predetermined number n.

On the other hand, in the i-th input image, the position (x,
If the input value to the pixel of y) is expressed as _i (x, y), then
First input value _i (x, y) and i=1 to i=n
The added value of the difference from the input value _i (x, y) up to is _o 〓 ⁱ⁼¹ { ₁ (x, y) − _i (x, y)} (1). Next, by subtracting equation (1) from n times _i (x, y), we get n ₁ (x, y)− _o 〓 ⁱ⁼¹ { ₁ (x, y)− _i (x, y)} = [ _o 〓 ⁱ⁼¹ _i (x, y)] (2). This value is nothing but the added value of images input n times. Here, ₁ (x, y) is stored in the first image memory, and _o 〓 ⁱ⁼¹ { ₁ (x, y) - _i (x, y)} is stored in the third image memory. Therefore, the first and third
By performing the calculation of equation (2) between image memories, a high-gradation added image can be obtained. Obtaining such highly gradated values has the advantage that even subtle changes in density can be recognized and measured.

Also, if both sides of equation (2) are divided by n, that is,
Subtracting the value of the third image memory divided by n from the first image memory yields ₁ (x, y) − [ _o 〓 ⁱ⁼¹ { ₁ (x, y) − _i (x, y)}] /n=[ _o 〓 ⁱ⁼¹ { ₁ (x, y)}]/n (3). This value is nothing but the average value corresponding to each pixel.

〔Effect of the invention〕

As explained above, focusing on the differences between images,
By inputting and storing a still image multiple times, it is possible to obtain high gradation or averaged image input values using normal image memory without using special image memory. It has the advantage of being economical.

[Brief explanation of the drawing]

FIG. 1 is a system diagram of an image processing apparatus used in the present invention, and FIG. 2 is a flowchart showing the method of the present invention. 1... Console CRT, 2... System processor, 3... Main memory, 4... ITV camera,
5... Inter-image arithmetic unit, 6... Image memory, 7...
Image processing processor, 8...data bus.

Claims (1)

[Claims] 1. Input and store the image from the image input device into a gradation image memory that can input and store gradation information for at least three planes, or add, subtract, or add weight between images in the image memory using an inter-image arithmetic unit. Square/
In a first step, an initial image from an image input device is input into a first image memory using an image processing device that has an image processing processor that performs division and controls this image processing processor according to a program by a system processor. Then, in a second step, the second and subsequent images from the image input device are input into a second image memory and sequentially updated and stored;
In the third step, the value in the second image memory is subtracted from the value in the first image memory by the inter-image arithmetic unit, and the difference is inputted to the third image memory and sequentially added and stored. An image input and storage method characterized in that up to three steps are repeated until the number of image inputs reaches a predetermined number.