JPH02257372A - Moving object identifying device - Google Patents

Abstract

PURPOSE:To correctly discriminate the form of a moving body by operating the difference between picture elements corresponding to each other of a present image and an initial image and distinguishing luminance levels of picture elements higher than a prescribed luminance level and those lower than it from each other to eliminate the background and preventing the image of a moving body from being taken into the background. CONSTITUTION:The initial image obtained by an image pickup means 1 is stored in a first area of an image memory part 2, and another image obtained next is stored in a second area of the image memory part 2. Luminance levels of picture elements on the same address of the present image obtained by the image pickup means 1 and the initial image stored in the first area of the image memory part 2 are compared with each other; and when the difference between them is larger than a threshold, the change of the luminance level is recognized to count up a picture element update counter 4 by +1. Luminance levels of picture elements on the same address of the preceding image and the present image are compared with each other by the luminance level comparing part 3; and when the luminance level is changed beyond the threshold, detection of the moving body is recognized to reset the picture element update counter. Thus, the occurrence of noise due to the change of the background or the change of luminance is suppressed to always identify the moving body.

Description

DETAILED DESCRIPTION OF THE INVENTION Field of the Invention The present invention relates to an apparatus for capturing a moving object in an image and identifying the moving object from the characteristics of the image.

Conventional technology Conventionally, this type of moving object identification device has generally used multiple infrared sensors to identify a moving object (hereinafter referred to as a human body).
There have been methods for detecting the temperature of the human body and detecting the position from the direction of the infrared sensor with the larger output (prior art example 1), and methods for capturing the human body in an image. Means for detecting a human body from this image include a means for extracting only the moving part by taking a difference screen between the background screen and the current screen (conventional example 2) as shown in FIG. Means for obtaining the so-called image differential value that takes the difference between the previous screen and the current screen (Conventional Example 3)
)was there.

Problems to be Solved by the Invention However, in Conventional Example 1, even if the approximate position is known, the accuracy is not high enough to recognize each part of a person, so complex control cannot be performed. there were.

In addition, in Conventional Example 2, since the background is further removed, the shape of the human body can be extracted and the position of each part can be detected from the accurate shape of the human body. If you take a difference with the screen, the areas where the background has moved will be extracted. (2) If the entire background shifts even slightly, noise will occur throughout the entire screen when the difference is taken from the current screen. (3) If the imaging means moves, noise will similarly occur on the entire screen.

(4) When there is a change in brightness (for example, when the brightness in the room changes), noise also occurs across the screen.

Furthermore, in Conventional Example 3, since the difference with the previous screen is taken, that is, it is a differential value of the image, (1), (
Issues like 2) will disappear, and issues like (3) and (4) will have less of an impact. However, (1) the edges of a moving human body are extracted, but not the entire human body, so when detecting the position, the edge lines may be broken, making detection difficult and resulting in incorrect positions being detected. Put it away. (2)
When there is less movement, the edge lines become thinner and eventually disappear, causing a problem in that the human body disappears and reappears, causing malfunctions.

Therefore, the present invention automatically changes the background by making corrections even if the background is moving, and by detecting the degree of subsequent change in the moving background. The purpose is to suppress noise caused by changes and to be able to identify only those who are constantly moving.

Means for Solving the Problems In order to achieve the above objects, the present invention includes an imaging means, an image memory section for storing a plurality of images obtained from the imaging means, and an image memory section that stores the same images in the image memory section. a brightness level comparison unit that compares the brightness levels of address pixels; a pixel update counter that represents the history of changes in the brightness level of the pixel based on the output of the brightness level comparison unit; and a continuity that determines the continuity of the pixel update counter. The apparatus includes a sentence determining section and a brightness level converting section for converting the brightness level of the image memory section using the continuous sentence determining section.

Operation With the above-described configuration, the moving object identification device of the present invention stores the initial image obtained by the imaging means in the first image in the image memory section, and similarly stores the next obtained image in the second image in the image memory section.
A brightness level comparing unit compares the brightness level between pixels at the same address of the current image obtained by the image pickup means and the initial image stored in the first image memory unit. , when the brightness level is greater than or equal to the threshold, it is assumed that a change in the brightness level has occurred, and the pixel update counter is incremented by one.

Here, the pixel update counter imports the current pixel into the initial pixel when the current pixel is successively different from the initial pixel. After that, the brightness level comparison unit compares the pixels at the same address in the previous image and the current image, and if there is a change in brightness level above the threshold, it is assumed that a moving object has been detected and the pixel update counter is reset. do. Then, the current pixel is taken into the second image memory unit, and the degree of continuity of the change counter is determined by the continuous sentence determination unit. If the continuity is greater than or equal to the setting, the current pixel is stored in the first image in the image memory section and imported into the initial image; otherwise, the brightness level is changed according to the brightness level of the current pixel and the initial pixel. Conversion is performed by the level conversion section. This allows the shape of the moving object to be determined. Then, it has the function of detecting, for example, a specific part of the previous image or current image after being converted in this way, and detecting the position of a moving object.

Example An embodiment of the present invention will be described below with reference to the accompanying drawings.

FIG. 1 is a block diagram of an embodiment of the present invention, in which 1 indicates an imaging means;
Reference numeral 2 denotes an image memory unit that stores the image inputted from the image pickup means 1; 3, a brightness level comparison unit that compares the brightness levels of pixels at the same address by performing inter-image calculations on the images stored in the image memory unit; 4 is the brightness level comparison section 3
A pixel update counter that indicates the history of changes in the brightness level of the pixel, 5 is a continuous sentence determination section that determines the continuity of the pixel update counter, and 6 is a continuous sentence determination section that determines the continuity of the pixel update counter. The brightness level converting section 7 converts the brightness level, and the position detecting section 7 detects a specific position in the image memory section 1 after the brightness level conversion section 6 partially or completely converts the brightness level. be.

Further, the image memory section 2, the brightness level comparison section 3, the change counter 4, the continuous sentence determination section 5, the brightness level conversion section 6, and the position detection section 7 are connected to the CPU by a bus 9.

Next, a description will be given based on the flowcharts of FIGS. 2 and 3. In step 101, initial settings such as resetting the image memory unit 2 and setting the comparison threshold value of the brightness level comparison unit 3 are performed.In step 102, an initial image is manually captured by the imaging means 1, and in step 103, the initial image is (hereinafter also referred to as a background image) is stored in the first image of the image memory section 2. Thereafter, in step 104, an image is inputted by the imaging means 1, and in step 105, it is stored in the second image in the image memory section 2. (The image stored here is hereinafter referred to as the previous image.) Next, in step 106, an image is input by the imaging means 1, and this image is used as the current image, and in step 107, the address of the pixel in the image is reset, and the address of the pixel in the image is reset to address 0. In step 103, the brightness level is compared with the 0th pixel of the initial image stored in the first image in the image memory section 2.
If the difference in the brightness level between the current pixel (representing one pixel of the current image) and the initial pixel (representing one pixel of the initial image), which is determined by the brightness level comparison unit 3 in step 8, is less than or equal to the threshold value determined in step 101. If so, the pixel update counter is reset in step 109, and if the difference in brightness level is equal to or greater than the threshold value, the lever is turned on, and the pixel update counter is incremented by 1 in step 110. Image 1
The brightness level of the current pixel (indicating one pixel) is compared with the brightness level of the current pixel in the brightness level comparison unit 3, and if it is equal to or higher than the threshold value, step 1
In step 12, the pixel update counter is reset, and if it is less than the threshold value, nothing is done, and in step 113, the current pixel is stored in the second sheet of the image memory section 2 at the same address and is replaced with the previous pixel. Replacement 0 Next, in step 114, the continuity of these luminance level changes is determined, and associated image processing is performed.In step 115, it is detected whether the determination of all pixels has been completed, and it is determined whether all the pixels have been determined. If not, the process returns to step 10B in order to amplify the pixel address and detect the next pixel in step 116, and if all is completed, the position detection section 7 detects a specific position in step 117, and the position detection section 7 detects the specific position in step 11B. The device is controlled by the output of 7.

Next, the flowchart of the continuity determination and image processing in step 114 will be explained based on FIG. The update counter 4 is reset, and the brightness level of the current pixel is set to O by the brightness level converter 6 in step 203. Also, if the continuity determining unit 5 determines that the pixel update counter 4 is continuous, it is determined in step 204 whether the degree of continuity has reached the maximum degree of continuity, and if the degree of continuity is not the maximum degree of continuity, in step 205 the pixel update value is The counter is reset, and in step 20G, the brightness level of the current pixel and the initial pixel are compared by the brightness level comparison unit 3. If the brightness level of the current pixel is higher than the set level, the brightness level of the current pixel is set to 1 in step 207, and if it is lower than the set level, the brightness level of the current pixel is set to 1. The process goes to step 203 and the brightness level of the current pixel is set to 0. Next, if the maximum degree of continuity has reached the maximum in step 204, the pixel update counter is incremented by 1 in step 208, and the upper limit of the repetition of the maximum degree of continuity is set at step 208.
9 to detect. If the repetition does not reach the upper limit, the process proceeds to step 206, where the brightness level of the current pixel and the initial pixel are compared by the brightness level comparison unit 3, and if the brightness level of the current pixel is equal to or higher than the set level, the brightness level of the current pixel is set to 1 in step 207, If it is below the set level, the process goes to step 203 and sets the brightness level of the current pixel to O. If the repetition of the maximum degree of continuity reaches the upper lIN, the pixel update counter 4 is reset in step 210, and the current pixel is stored in the image memory section 2 in step 211.
After steps 203, 207, and 211, in which images are captured at the same address in the first initial image, the process proceeds to step 115 in Figure 2, which asks whether all pixels have been detected. .

To explain in detail the changes in pixel brightness level shown in the above flowchart using Figure 4, Figure 4 shows whether or not four consecutive screens have changed with respect to the initial image. There are model B, background C, noise D, and noise E caused by changes in the background.The first stage is the initial image (image O), and the second stage is four consecutive images. 1 to image 4, and the third row represents the subtraction of image 1 to image 4 in the upper row. In addition, the following table shows that for one pixel in each image from image 0 to image 4, the initial pixel of image O is O, the pixel where the brightness level has changed by more than the threshold value is 1, and the pixel where the brightness level has changed by less than the threshold value is 1. Place O for those that did not change or did not change at all.
The types of change patterns for pixels of four images from image 1 to image 4 are shown.

In other words, for pattern l, the contents of the pixel update counter are oo
oo, where there is no change at all with respect to the initial pixel, which corresponds to the background C and desk A in Figure 4, and patterns 2 to 15 indicate moving objects, and in this case, the D Pattern 10 corresponding to the case and B shown as a human body model also show one of these patterns. Pattern 16 is noise that appears after the background, and is a case where the content of the pixel update counter changes to 1111 four times in a row with respect to the initial pixel. It is incorporated into the background by repeating the image once or by repeating it a predetermined number of times. As a result, D eventually ceases to be noise and enters the initial image as a new background, and thereafter only the human body, which is a true moving object, is extracted.

Effects of the Invention As described above, the present invention calculates the difference between the same pixels in the current image and the initial image, and distinguishes between the brightness level of pixels above a predetermined brightness level and the brightness level below (other than 1 and O). For pixels that always repeat the same change with respect to the background, the continuity and repetition of continuity are detected and incorporated into the background. By this, it is possible to respond to changes in the background, detect changes between the previous pixel and the current pixel, and if there is a change of more than a predetermined brightness level, cancel the continuity described above and prevent the moving object from being unnecessarily added to the background. It is designed to prevent this from occurring, and has the following effects:

(1) Since the difference between the current pixel and the initial pixel is performed, it is easy to see the shape of the moving object on the screen where all pixels are combined (this is a feature of the method of difference from the background)
Then, if this background changes by the same amount within a certain amount, only that pixel is incorporated into the background, so that it is possible to correct the change in the background. Therefore, even if the background shifts or the imaging means moves, the background is rewritten, so only the human body can be extracted at all times.

(2) In addition to the difference between the background and the current pixel, the difference between the previous pixel and the current pixel is performed, that is, the instantaneous value of the image is detected, so moving parts are reliably detected as moving objects.

(3) Since inter-image calculations are set for each pixel by a pixel update counter according to its characteristics, detailed corrections can be made.

[Brief explanation of drawings]

Fig. 1 is a configuration diagram of a moving object identification device according to an embodiment of the present invention, Fig. 2 is an overall flowchart of the same, Fig. 3 is a flowchart of the continuity determination section, and Fig. 4 is a model explanatory diagram of the same continuity. , FIG. 5, and FIG. 6 are diagrams showing images of conventional examples. 1... Image means, 2... Image memory unit, 3... Brightness level comparison unit, 4... Change counter, 5...・Continuity determination section, 6...
...Brightness level conversion section. Name of agent: Patent attorney Shigetaka Awano (1 person)

Claims (1)

[Claims] an image capturing means, an image memory section for storing a plurality of images obtained from the image capturing means, a luminance level comparison section for comparing the luminance levels of pixels at the same address between the plurality of images in the image memory section; A pixel update counter that represents the history of changes in the brightness level of pixels based on the output of the brightness level comparison section, a continuity determination section that determines continuity of the pixel update counter, and a continuity determination section that controls the image memory section. A moving object identification device comprising: a brightness level changing unit that converts the brightness level of a moving object;