JPH01273198A - Image monitoring device - Google Patents

Abstract

PURPOSE:To detect whether or not there is an intruder without any malfunction by judging whether or not the intruder, etc., is present by using a threshold value corresponding to the area of each monitor area divided for the output of an area-classified binarizing means. CONSTITUTION:An area-classified variation detecting means 7 finds the difference in brightness value between reference images from a current image storage means 3 and a reference image storage means 6, calculates the sum of the absolute values of difference of respective areas by using stored division area information and threshold values by the areas, and compares the sum with the threshold values by the areas to detect variation. An area-classified mean value arithmetic means 8 finds the mean values of current images by the areas where there is variation and an area-classified mean value subtracting means 9 subtracts the mean values from the brightness values of the current value by the areas. Then an area-classified binarizing means 10 converts the subtraction results into binary values to extract effective picture elements and a decision means 11 compares the number of the effective picture elements with the threshold values by the areas to decide whether or not there is an intruder. Consequently, the presence/absence state of the intruder is detected without any malfunction.

Description

[Detailed Description of the Invention] [Industrial Application Field] The present invention provides an image capturing device such as a television camera to detect the presence or absence of an intruder by comparing the current image with a reference image. This invention relates to a monitoring device.

[Conventional technology]

A conventional device of this type is shown in FIG.

The image signal from the imaging means l such as a television camera is A/D
The difference image storage means 12 generates a difference image between the current image sampled and digitized by the conversion means 2 and stored in the current image storage means 3 and a reference image such as a background image stored in advance in the reference image storage means 6. 2 (The unification means 13 separates the brightness value of each pixel of this difference image into invalid data and valid data using a predetermined darkness value, and the determination means 11 determines the brightness value of the valid data. The idea was to detect intruders into the monitoring area based on changes in the number (Japanese Patent Laid-Open No. 160282/1982).

However, in such a device, if a change in brightness occurs in a certain area of the monitoring area due to a change in sunlight, the difference in brightness from the reference image will change to a certain darkness even if there is no abnormality such as the presence of an intruder. The problem is that the number of pixels exceeding the specified value will exceed the predetermined value and issue a false alarm, so it is not practical in places where the illuminance changes, such as near windows, and can only be used indoors with stable lighting. There was a point.

In order to improve the above-mentioned problems, a method has been devised in which the reference image is updated at regular intervals to cope with changes in illuminance.

[Problem to be solved by the invention]

However, this method cannot handle cases where the illuminance changes quickly compared to the update time of the reference image, and there is still a problem that it is still not possible to sufficiently identify intruders into the monitoring area and illuminance changes. Ta.

The present invention has been made in view of the above points, and its purpose is to provide an image that can detect the presence or absence of an intruder without malfunctioning even in the event of sudden changes in illuminance within a monitoring area. The objective is to provide a monitoring device.

[Means to solve the problem]

The image monitoring device of the present invention includes: an imaging means for photographing a monitoring area; an A/D conversion means for sampling and A/D converting an image signal output from the imaging means;
current image storage means for sequentially storing the current image output from the D conversion means; reference image storage means for storing a reference image representing the normal state of the monitoring area; A region dividing means that divides a region based on correlation and sets a threshold according to the area of each divided region; a region division information storage means that stores the region division information and the threshold; and a brightness of the current image for each region. area-based average value calculation means for calculating the average value of the values; area-based average value subtraction means for subtracting the average value from the brightness value of the current image for each area; and a determination means for determining whether there is an intruder or the like based on the darkness value of the output of the region-specific binarization means. It is.

[Effect]

The area dividing means divides the monitoring area into areas with similar correlations in temporal changes in brightness, calculates the average value of brightness for each area for the difference image between the current image and the reference image, and calculates the average value of the brightness for each area. By subtracting the average value from the luminance value and binarizing the result for each region, effective pixels are obtained, and by comparing the number with a threshold proportional to the area of each region,
Determine whether there is an intruder, etc.

〔Example〕

Hereinafter, one embodiment of the present invention will be described based on the drawings.

FIG. 1 is a block diagram showing one embodiment of the present invention.

Reference numeral 1 denotes an imaging means such as a television camera, which takes an image of the monitoring area and outputs an image signal. 2 is an A/D conversion means,
The image signal is sampled and digitized. Reference numeral 3 denotes a current image storage means for sequentially storing the digitized current image output from the A/D conversion means 2.

Reference numeral 4 denotes a region dividing means which divides the monitoring region into small regions having similar correlations in time change in brightness of each pixel. That is, for example, the monitoring area is monitored for about one day in advance, the changes in the brightness value of each pixel over the course of the day are checked, and areas that change in the same way are divided. Furthermore, it has a function of determining a threshold value proportional to the area of each divided region. There are two types of threshold values: one for brightness and the other for the number of pixels; the former is used as a darkness value in the region-by-region change detection means described later, and the latter is used as a threshold value in the determination means described later.

Reference numeral 5 denotes a region division information storage means, which stores information on the divided regions obtained by the region division means 4 and threshold values for each region (
, 6 is a reference image storage means for storing reference images representing the normal state of the monitoring area.

Reference numeral 7 denotes a change detection means for each region, which calculates the difference in luminance value between the current image from the current image storage means 3 and the reference image from the reference image storage means 6, and calculates the difference in luminance value between the current image from the current image storage means 3 and the reference image from the reference image storage means 6, and calculates the difference in the luminance value from the current image from the current image storage means 3 and the divided region information stored in the region division information storage means 5. By using the threshold value for each region and the sum of the absolute values of the differences for each region, and comparing it with the threshold value for each region,
Perform change detection.

Reference numeral 8 denotes an average value calculation means for each area, which calculates the average value of the brightness of the current image for each area for the area where the change detection means 7 for each area has changed.

Reference numeral 9 denotes an average value subtracting means for each area, which subtracts the average value from the brightness value of the current image for each area.

IO is a region-by-area binarization means, which binarizes the subtraction result of the region-by-area average value subtraction means 9 and extracts effective pixels.

Determining means 11 determines whether or not there is an intruder by comparing the number of effective pixels obtained by the area-based binarization means with the threshold value for each area.

Next, the operation of this embodiment will be explained.

The image signal obtained by the imaging means 1 is converted into a digital signal by the A/D conversion means 2 and stored in the current image storage means 3. The area dividing means 4 divides the monitoring area into small areas having similar correlations in temporal changes in brightness of each pixel, and calculates a darkness value proportional to the area of each area.This information is stored in the area dividing information storage means. The 0wI area change detection means stored in the reference image storage means 5 calculates the difference in the absolute value of the luminance value between the reference image stored in the reference image storage means 6 in advance and the current image, and calculates the sum of the difference values for each area. When the threshold value is exceeded, it is determined that there is a change.

The area-specific average value calculation means 8 calculates the average value of the brightness values of the current image for the area where there has been a change, and the area-specific average value subtraction means 9 subtracts this average value from the brightness value of the current image. . The result of this subtraction is binarized by the area-based binarization means 10, and it is determined whether or not there is an intruder based on the determined number of valid pixels using the area-based threshold value.

Here, for each region with similar correlation of time changes in brightness values,
Using the result of subtracting the average brightness value for each area,
Since the presence of an intruder is determined by converting the value into a value, if the illuminance changes due to a change in illumination, etc., the illuminance distribution will be uniform in each area, and if the average value is subtracted, it will become approximately zero. Therefore, even if this subtraction result is binarized, there are almost no effective pixels.

On the other hand, if there is an intruder, even if you subtract the average value,
Since it does not become zero at the intruder's location, it becomes an effective pixel when disulfide occurs.

Note that in the present invention, the region-by-region change detection means 7 can be omitted, and in this case, the region-by-region average value calculation means 8 only needs to calculate the difference between the current image and the reference image, and the region division information etc. can be directly transmitted. It is input to the average value calculating means 8 for each region.

However, in the case where the region-by-region change detection means 7 is provided, the amount of calculation can be reduced because it is only necessary to calculate and subtract the average value in the region where a change has occurred.

〔Effect of the invention〕

The image monitoring device of the present invention includes: an imaging means for photographing a monitoring area; an A/D conversion means for sampling and A/D converting an image signal output from the imaging means;
current image storage means for sequentially storing the current image output from the D conversion means; reference image storage means for storing a reference image representing the normal state of the monitoring area; an area dividing means for dividing an area based on correlation and setting a threshold value according to the area of each divided area; an area dividing information storage means for storing the area dividing information and a darkness value; A region-by-region average value calculation means for calculating the average value of brightness values, a region-by-region average value subtraction means for subtracting the average value from the brightness value of the current image for each region, and an output of the region-by-region average value subtraction means. Since it is made up of an area-specific binarization means that binarizes each area, and a determination means that judges the presence or absence of an intruder based on the threshold value with respect to the output of the area-specific binarization means, the monitoring area An image monitoring device capable of detecting the presence or absence of an intruder without malfunctioning even if there is a sudden change in illumination inside the room has been provided.

[Brief explanation of the drawing]

FIG. 1 is a block diagram showing one embodiment of the present invention, and FIG. 2 is a block diagram showing a conventional example. 1--Imaging means 2--A/D conversion means 3--Current image storage means 4--Region division means 5--Region division information storage means 6-- Reference image storage means 7 - Area change detection means 8 - Area average value calculation means 9 - Area average value subtraction means 10 - Area binarization means 11 - Judgment means Patent applicant Matsushita Denko Co., Ltd. Agent Patent Attorney Takemitsu Sanko (and 2 others) Figure 1

Claims (1)

[Claims] (1) Imaging means for photographing the monitoring area, sampling the image signal output from the imaging means, and A/D
A/D conversion means for converting, current image storage means for sequentially storing current images output from the A/D conversion means, and reference image storage means for storing a reference image representing a normal state of the monitoring area; area dividing means that divides the monitoring area based on the correlation of temporal changes in brightness of each pixel and sets a threshold value according to the area of each divided area; and area dividing information that stores the area dividing information and the threshold value; an accumulation means, an average value calculation means for each area for calculating an average value of the brightness values of the current image for each area, an average value subtraction means for each area for subtracting the average value from the brightness value of the current image for each area; region-specific binarization means for binarizing the output of the region-specific average value subtraction means for each region; and determination for determining the presence or absence of an intruder based on the threshold value for the output of the region-specific binarization means. An image monitoring device comprising means.