JPH04235489A - Image monitoring device - Google Patents

Abstract

PURPOSE:To decide the change of the brightness of a back ground such as the fluctuation of illumination by distinguishing it from an intruding object. CONSTITUTION:This image monitoring device is provided with a measuring part 2 to find the number of picture elements with the same brightness in the image picked up by a image pickup part 1 corresponding to the respective brightness of the picture elements, a detecting part 3 to determine the difference between the images of the number of the picture elements and a deciding part 4 to decide the presence of the intruding object when the difference is prominently large within the limits of the fixed brightness. Further, when periodic fluctuation in the back ground occurs, the image is picked up when the phase on the period of the fluctuation is the same by providing an image pickup control part in front of the image pickup part 1.

Description

[Detailed description of the invention]

[Purpose of the invention]

0001

[Industrial Application Field] The present invention is applicable to banks, department stores,
The present invention relates to an image monitoring device used to detect intruders, intruding objects, suspicious persons, etc. in buildings, etc.

0002

BACKGROUND OF THE INVENTION In recent years, with the development of the information society, the importance of various types of security has increased. As part of this,
There are image monitoring systems used to detect intruders, intruding objects, suspicious persons, etc. In the past, images on monitors were monitored by humans, but as the number of areas to be monitored increases, the burden of constant monitoring on monitors increases and reliability decreases, making automation highly desirable.

[0003] Recently, several systems have been proposed that perform automatic monitoring using TV cameras. The method used for these is to take the direct difference between images obtained in time series or between the background image, calculate the total amount of change within a preset area, and combine this total amount with a threshold value. It is common to detect and determine abnormalities by comparing the

However, in such a conventional monitoring system, in an environment where the illumination intensity fluctuates, such as outdoors or under fluorescent lighting, changes in the brightness of the background area due to illumination fluctuations occur, making it difficult to intrude. Separate changes caused by objects,
There was a problem in that it was difficult to detect and the monitoring accuracy decreased. In particular, the effects of illumination fluctuations tend to be noticeable, such as on the edges of objects in the background. Although false detections frequently occurred, no measures were taken to eliminate the effects of illumination fluctuations.

[0005]

[Problem to be Solved by the Invention] In this way, conventional image monitoring devices perform processing that takes direct differences between images and calculates the total amount of change. Even if there were, the imaging was performed asynchronously with the illumination, which caused a problem in that the detection performance deteriorated significantly in an environment where the illuminance of the illumination fluctuated.

The present invention has been made in consideration of the above circumstances, and its object is to provide an image monitoring device that operates stably in an environment where the illumination intensity fluctuates. [Structure of the invention]

[0007]

[Means for Solving the Problems] An image monitoring device according to the present invention includes a measuring means for determining the number of pixels having the same brightness in a captured first image for each brightness, and a measuring means for determining the number of pixels having the same brightness in a first image taken. detection means for determining, for each brightness, the difference between the number of pixels of the first image taken and the number of pixels found by the measuring means of a second image taken before the first image; determination means for determining an abnormality in the first image based on the difference determined by the detection means;

000
8] The determination means determines, for example, if the variance of the difference determined by the detection means is greater than or equal to a predetermined value, or if the difference between the maximum and minimum values of the differences determined by the detection means is greater than or equal to a predetermined value. If it is, it is determined to be abnormal.

[0009] In addition, in an image monitoring device that detects changes in brightness from captured images and determines abnormalities based on this change in brightness, when there is a periodic change in the imaging location, it is possible to detect changes in this change. The present invention is characterized by the addition of a period measuring means for measuring the period, and an imaging control means for controlling the imaging to be performed at a time point on the period measured by the period measuring means when the phases are the same.

0010

[Operation] In the present invention, images are taken at any time with a TV camera, and the difference between the currently captured image and a reference image, which is the previously captured image or a previously captured background image, is checked. Performs intrusion detection. The brightness distribution (density histogram) in the image is examined as an index used for intrusion detection. This is to find the number of pixels having the same brightness (density value) in the image for each brightness (density value). Then, the difference in the number of pixels having the same brightness (density value) is determined between the current image and the reference image. Intrusion detection is performed based on the results of calculating this difference for each brightness (density value). That is, if the difference is significantly larger than other parts within a certain range of concentration values, it is determined that there is an abnormality (intruder). As a result, when there is only a variation in illumination, for example, the difference appears almost equally across all density values, and it is determined that there is no abnormality (intruder), so that the variation in illumination is no longer mistakenly detected as an intruder.

[0011] Furthermore, if there are periodic fluctuations in the imaging location (including fluctuations in illumination), the brightness of the background portion can be changed by controlling the imaging to be performed at a point in time when the phase on this period is the same. I don't get the image. By controlling imaging in this manner, false detections can be reduced even with conventional image monitoring devices that use the sum of direct differences between images as an index for intrusion detection.

0012

DESCRIPTION OF THE PREFERRED EMBODIMENTS An image monitoring apparatus according to an embodiment of the present invention will be described below with reference to the drawings. ○Example 1

001
3] FIG. 1 is a schematic configuration diagram of an image monitoring device according to this embodiment,
FIG. 2 is a flow diagram showing an algorithm for image monitoring according to this embodiment. In FIG. 1, an imaging unit 1 sequentially captures time-series images taken by a TV camera into an image memory of a processing device. If a background image without any intruders is used as a reference image, this is also imported (FIG. 2S1). Next, measurement section 2
Now, by scanning the image memory described above, the number of pixels having the same density value (this is called brightness distribution data) is counted and stored (S2 in FIG. 2). Then, the detection unit 3 calculates the difference between the images of the brightness distribution data stored above (S3 in FIG. 2), and the determination unit 4 makes an abnormality determination based on this difference. In other words, if a difference occurs (Figure 2S4)
, we look at whether this difference stands out in a certain density value range (Figure 2S5) and determine whether this difference is truly due to an intruder (Figure 2S6) or whether it is due to changes in background brightness due to illumination fluctuations, etc. It is determined whether there is any invading object (S7 in FIG. 2). The details of each part in FIG. 1 will be described below.

[0014] In the imaging unit 1, a TV set in a monitoring environment is
The video signal output from the camera 5 is sent to an NTSC decoder 6,
The images are digitized via the A/D converter 7 and stored as time-series images in the image memory group 9 of the device via the ITV interface 8.

In the measuring section 2, the pixel number counter 12 prepared for each quantized density value is cleared in advance, and the image coordinate specifying circuit 10 scans the readout position of the image memory for measuring the brightness distribution. Then, one pixel at a time is taken out and input to the density measurement circuit 11. The density measuring circuit 11 repeats the process of measuring the density of the input pixel and adding one to the corresponding pixel number counter. When the scanning of one image is completed, the contents of the pixel number counter are transferred to the brightness distribution data memory group 13 and stored therein. When the density of each pixel is quantized using 8 bits, brightness distribution data is obtained as an array consisting of 256 elements, as shown in the graph of FIG. 4, for example.

In the detection section 3, the difference circuit 14 calculates the absolute value difference between each element of the array stored in the brightness distribution data memory 13. In the determination unit 4, the sum of the absolute value differences is calculated in parallel by a sum calculation circuit 15, and the variance value is calculated in parallel by a variance value calculation circuit 16, and inputted to an intrusion determination circuit 17. In the intrusion determination circuit 17, the sum of the absolute value differences is a predetermined threshold value (A).
If it is larger (FIG. 2S4Yes), the variance value is compared with a predetermined threshold value (B) (FIG. 2S5). The predetermined threshold value (A) is determined regarding the summation, and the predetermined threshold value (B) is determined regarding the variance. If the variance value is smaller than the predetermined threshold value (B), it is determined that the change is due to illumination variation (S7 in FIG. 2). Default threshold (B)
If it is larger, it is determined that it is caused by an invading object (Figure 2
S6). The above-mentioned variance value calculation circuit 16 is replaced with a (maximum value - minimum value) calculation circuit, and the predetermined threshold value (B) is replaced with one determined based on the difference between the maximum value and the minimum value, and the determination is made in the same manner. It's okay.

[0017] When there are no intruders and only illumination fluctuations occur, the shape (envelope) of the brightness distribution does not change significantly, as shown in Figure 5(a), and when the illumination becomes brighter, the shape of the brightness distribution changes to the right as a whole. When the illumination becomes dark, there is a tendency for the overall density to shift only to the left (towards the lower density value). Therefore, the absolute value difference between the two determined by the detection unit 3 is a substantially uniform value over the entire density value range, as shown in (b). Here, when the dispersion or the difference between the maximum value and the minimum value of this absolute value difference is calculated in the determination unit 4, it becomes a small value.

On the other hand, if an intrusion occurs, FIG.
As shown in a), the shape of the brightness distribution in the concentration range of the intruder changes greatly. Therefore, the absolute value difference between the two determined by the detection unit 3 takes a particularly large value in the concentration range of the intruder, as shown in (b). Here, when the dispersion or the difference between the maximum value and the minimum value of this absolute value difference is taken in the determination unit 4, it becomes a large value, so it can be detected separately from the case of only illumination fluctuation.

In the above embodiment, the range for determining the brightness distribution is the entire image, but it can be modified to speed up the processing speed by limiting it to a certain area by controlling the image coordinate specifying circuit 10. It is also possible to use it as In addition, when the possible concentration range of the intruder is known in advance, the difference circuit 14, the sum calculation circuit 15, and the variance value calculation circuit 1
It is also possible to use a modification to increase the processing speed by limiting the data processed in step 6 to this density range. As described above, the image monitoring device according to the first embodiment is capable of distinguishing changes in background brightness due to illumination fluctuations and the like from intruders, regardless of the presence or absence of periodicity. ○Example 2

FIG. 3 is a schematic diagram of the image monitoring apparatus according to this embodiment. The imaging control unit 18 performs control so that imaging is performed at a point in time when the periodic phase of the illumination illuminance change measured using the illumination meter 21 or the like is the same. The imaging unit 19 sequentially captures time-series images taken by the TV camera under the control of the imaging control unit 18 into the image memory of the processing device. If a background image with no intruders is used as a reference image, this is also imported. The intrusion detection unit 20 determines whether an intruder is present by analyzing changes in the time-series images stored in the image memory.

FIG. 7 shows how the illumination illuminance changes periodically. t0, t1, t2... are the points when the illumination intensity reaches its maximum, and this interval is the period. For example, in the case of a fluorescent lamp using a normal ballast, this period is the AC power frequency (50
Hz/60Hz). However, this embodiment can be applied not only to illumination fluctuations but also to periodic fluctuations. For example, if there is a clock pendulum or metronome hand that moves periodically in the background (although the frequency in this case is very small compared to the AC power frequency), Figure 7
You can think of t0, t1, t2... as the point in time when a pendulum or needle comes to the center. The details of each part in FIG. 9 will be explained below.

In the imaging control section 18, a timer 22 measures time, and a phase control circuit 23 determines the point in time when the phase of a change in illumination intensity or periodic fluctuation of the background is the same. The shutter opening/closing control circuit 24 controls the shutter attached to the TV camera 5 so that the shutter is opened for a preset time length of 2T around each of these points in time. The time length 2T is set sufficiently short with respect to the cycle. For example, if it is set to perform imaging in synchronization with the point in time when the illumination is the brightest, the above-mentioned t0, t1,
The shutter opens in the time range hatched around t2.... The reason why the image is taken at the point in the cycle when the illuminance is the highest is that the dynamic range for determining whether or not an intruder is present can be widened.
. . . may be set by the user before use, such as when the phase is π/4.

Due to the function of the imaging control unit 18 described above, periodic background fluctuations (such as illumination fluctuations and movement of objects other than the target object) no longer appear in the time-series images, so that these are erroneously detected as intruders. This can be prevented from happening. The configuration of the imaging section 19 is similar to that shown in FIG.

In the intrusion detection section 20, a difference circuit 25 calculates the difference between the images stored in the image memory group 9, and a sum calculation circuit 26 calculates the sum of these differences. The threshold comparison circuit 27 compares this sum with a predetermined threshold, and if the sum is larger, it is determined that there is an intruder, and the alarm device 28 issues an alarm, for example.

[0025] In the intrusion detection section 20 described above, the sum of direct differences is taken as an index used for intrusion detection.
By using the measuring section 2, the detecting section 3, and the determining section 4 in the first embodiment instead of the intrusion detecting section 20, the index used for intrusion detection can be the variance of the difference in brightness distribution data. In other words, the imaging control section 18 of the second embodiment is provided in front of the imaging section 1 of the first embodiment. In this way, if there is periodicity in the illumination fluctuations, the influence of the illumination fluctuations can be cut off before imaging, so the absolute value difference in Figure 5(b) becomes almost zero, indicating that there is an intruder. 6(b) becomes clearer in a binary manner, and the frequency of false detections can be further reduced.

In the above embodiment, it has been described that intrusion is determined based on the difference/sum of the entire image, but the sum calculation circuit 26
It is also possible to use a modified version to shorten the processing time by limiting the calculation area. Above, Example 2
The image monitoring device according to the above is capable of eliminating periodic background fluctuations before imaging.

[0027]

[Effects of the Invention] As explained above, according to the present invention, it is possible to reduce the possibility of falsely detecting changes in brightness due to illumination fluctuations and periodic fluctuations in the background as being caused by an intruder, thereby improving reliability. This brings about great practical effects, such as being able to provide an image monitoring device that can monitor images.

[Brief explanation of the drawing]

FIG. 1 is a schematic configuration diagram of an image monitoring device according to a first embodiment.

FIG. 2 is a flow diagram showing an algorithm of the image monitoring device according to the first embodiment.

FIG. 3 is a schematic configuration diagram of an image monitoring device according to a second embodiment.

FIG. 4 is a diagram showing an example of brightness distribution data.

FIG. 5 is a diagram showing an example of changes in brightness distribution due to illumination fluctuations.

FIG. 6 is a diagram showing an example of a change in brightness distribution due to an intruder.

FIG. 7 is a diagram showing periodic changes in illumination intensity and shutter opening time.

[Explanation of symbols]

1...Imaging unit 2...Measuring part 3...Detection section 4... Judgment section 5...TV camera 6...NTSC decoder 7...A/D converter 8...ITV interface 9...Image memory group 10...Image coordinate designation circuit 11...Concentration measurement circuit 12...Pixel number counter 13...Brightness distribution data memory group 14...Differential circuit 15...Sum calculation circuit 16...Dispersion value calculation circuit 17...Intrusion determination circuit 18...Imaging control unit 19...Imaging section 20...Intrusion detection unit 21...Luminance meter 22...Timer 23...Phase control circuit 24...Shutter opening/closing control circuit 25...Differential circuit 26...Sum calculation circuit 27...Threshold comparison circuit 28...Alarm circuit

Claims (5)

[Claims] 1. Measuring means for measuring the number of pixels having the same brightness in a captured first image for each brightness; a detection means for detecting, for each brightness, a difference between the number of pixels of a second image captured before the first image and the number of pixels determined by the measurement means; 1. An image monitoring device comprising: determination means for determining an abnormality in one image. 2. The image monitoring system according to claim 1, wherein the determining means calculates the variance of the difference determined by the detecting means, and determines that the image monitoring apparatus is abnormal if the variance is equal to or greater than a predetermined value. Device. 3. The determination means calculates the difference between the maximum value and the minimum value of the differences determined by the detection means, and determines that the difference is abnormal if the difference between the maximum value and the minimum value is greater than or equal to a predetermined value. The image monitoring device according to claim 1, wherein the image monitoring device makes a determination. 4. The image monitoring device according to claim 1, further comprising a period measuring means for measuring the period of the fluctuation when there is a periodic fluctuation in the imaging location; 1. An image monitoring device further comprising an imaging control means for controlling imaging to be performed at a point in time when the phases of the images are the same. 5. In an image monitoring device that detects a change in brightness from an image captured by a camera and determines an abnormality based on this change in brightness, when there is a periodic fluctuation in the imaging location, An image characterized by adding a period measuring means for measuring the period of fluctuation, and an imaging control means for controlling the camera to take an image at a point in time when the phases on the period measured by the period measuring means are the same. Monitoring equipment.