JP2923652B2 - Monitoring system - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION <Industrial Application Field> The present invention relates to a moving object detection device, and more particularly, to a moving object detection device based on a change in luminance of an image captured by a two-dimensional image sensor such as a television camera at predetermined time intervals. The present invention relates to a moving object detection device for detecting.

<Conventional technology> Various types of moving object detection devices are used in security and surveillance systems. Conventionally, one-dimensional sensors such as infrared photoelectric sensors, loop switches, and vibration switches are mainly used. The detection area of the sensor is small, and only a moving object can be detected in a narrow primary or two-dimensional area on a point or a line.

By detecting a moving object in a three-dimensional wide area, which was difficult with such a one-dimensional sensor, a sensor of a monitoring system that performs security and monitoring over a wide area more closely to the situation monitored by human eyes As a method, a method of specifying a moving object from a captured image using a two-dimensional image sensor such as a television camera has been considered.

However, in this type of moving object detection, a manned system in which an observer constantly monitors an image captured by a television camera to find a moving object is used, or such an image is temporarily recorded by a video tape recorder. This is a system for finding a moving object later from a reproduced image. Here, the image may be a moving image, or may be a still image at predetermined time intervals in consideration of economy,
In either case, the presence or absence of a moving object is ultimately determined by human eyes.

For this reason, especially when detecting the presence or absence of a moving object in real time, it is necessary to constantly monitor the monitor, which increases the burden on the observer, especially when simultaneously monitoring the monitors of cameras provided at a plurality of locations. Could overlook the presence of moving objects such as intruders.

As a moving object detecting device that solves such a problem, a current image signal captured by a two-dimensional image sensor is compared with a signal obtained a predetermined time before, and a moving object is detected based on a luminance difference between two image signals. A device for automatically detecting an object has been proposed (see Japanese Patent Application Laid-Open No. 64-48686).

That is, when there is no moving object in the captured image, an image that is substantially the same as the image at the present time and the image before the predetermined time should be obtained, and a large luminance change does not occur. When a person enters the camera, a large luminance difference occurs between the image data before and after the person enters, and the presence or absence of a moving object is detected by capturing the occurrence of the luminance difference.

<Problem to be Solved by the Invention> By the way, in the unmanned monitoring system using the two-dimensional imaging sensor as described above, when a moving object is detected based on the luminance difference output, an alarm is issued to the outside each time. When detecting the intrusion of an object into the monitoring field of view, an alarm should be output to the outside every time a moving object is detected based on the luminance difference output as described above, but it moves within the monitoring field of view. In the detection of an object, a warning may be output even when the object crosses the corner of the field of view, thereby causing a false alarm.

That is, when it is desired to detect an object that moves in the monitoring field of view, if an alarm is issued for each detection based on the luminance difference output, an object that moves in the field of view is detected, or it crosses the corner of the field of view. It is not possible to judge whether the detected object is a detected object, and the processing conditions (detection sensitivity, etc.) for the luminance difference output
The above-mentioned determination cannot be performed with high accuracy even if is set variously.

In addition, when a moving object is detected based on the luminance difference output, an alarm is issued to the outside each time a moving object is detected. -If the surrounding brightness changes in a short time due to turning off, there is a concern that this may be interpreted as a brightness change due to a moving object and a false report may be output.

The present invention has been made in view of the above problems, and by improving the characteristics of an alarm that has been output to the outside each time a moving object is detected based on a luminance difference output, to prevent false reports at the time of a single manifestation image. And improve the reliability of the monitoring system.

<Means for Solving the Problems> Therefore, the present invention includes two-dimensional imaging means for imaging a monitoring location, and the image signal at the present time and the image signal at a time earlier by a predetermined time imaged by the two-dimensional imaging means. In a monitoring system configured to detect a moving object in an image captured based on the luminance difference output by comparing the moving object detection based on the luminance difference output, the moving object detection based on the luminance difference output is continuously performed for a predetermined number of times or more. An alarm is issued to the outside only when it is done.

<Operation> According to this configuration, the current signal of the image signal captured by the two-dimensional image capturing means and the signal of a predetermined time earlier are compared, and a luminance difference output is obtained. Here, a moving object in the captured image is detected based on the luminance difference output, and an alarm is issued to the outside only when the number of continuous detections of the moving object based on the comparison between the two images is equal to or more than a predetermined number. If the number of continuous detections of the moving object is less than the predetermined number, no alarm is issued even if the moving object is detected, and no alarm is issued for the one-time detection of the moving object.

<Examples> Examples of the present invention will be described below.

FIG. 1 shows an embodiment of a monitoring system for alarming the presence of a moving object at a specific monitoring location to the outside. The monitoring system is a two-dimensional imaging means including a solid-state imaging device using a CCD or the like. The dimensional imaging sensor 1 is installed toward the specific monitoring location, and captures the monitoring location and outputs an analog image signal.

The analog image signal output from the two-dimensional image sensor 1 is clamped by the clamp circuit 2 while the A / D converter 3
Is converted into a digital image signal, and the luminance is quantized into, for example, 256 gradations represented by 8 bits.

The latest (current) digital image signal B is compared with the image A at a point in time before the predetermined time ΔT stored in the video memory 4 for each pixel, and a gradation difference C of luminance for each pixel is obtained.
(Luminance difference output) is calculated. Here, assuming that a pixel is represented by X and Y coordinates, the gradation difference (luminance difference) C is represented by C
(X, Y) = | A (X, Y) -B (X, Y) |

When the image stored in the video memory 4 is compared with the latest image, the latest image is used as a new stored image (reference image).
Is stored in the video memory 4 as
The image stored in the video memory 4 is updated every predetermined time ΔT.

When the gradation difference C (X, Y) in pixel units between the latest image (comparison image) and the image (reference image) at a point in time before the predetermined time ΔT stored in the video memory 4 is calculated, it can be arbitrarily set externally. The gradation difference j which is a threshold value of the luminance difference to be set is compared with the gradation difference C (X, Y) for each pixel, and only the pixels whose luminance has changed more than the gradation difference j are selected. Data D (X, Y) in which all pixels are binarized by dividing into pixels having changed luminance (1 is set as data) and pixels not changing luminance (0 is set as data) And Therefore, even if the luminance changes during the predetermined time ΔT, a pixel whose change is smaller than the gradation difference j is treated as if there is no luminance change, and natural phenomena and other noises of the video signal and the like are treated. This is intended to prevent erroneous detection of a luminance change unrelated to the moving object.
Thereby, the sensitivity of the moving object can be adjusted.

When all the pixels are binarized according to the presence or absence of the luminance change within the predetermined time ΔT as described above, the mask area (mask area map) stored in the mask memory 5 which is also arbitrarily set from the outside. M) invalidates the luminance change detection in the area not to be monitored. Specifically, for example, a pixel having a luminance change is set to 1, a pixel having no gradation change is set to 0, and a mask pixel not to be monitored is set to 1, and a non-mask pixel to be monitored is set to zero. In this case, the logical change of M (X, Y) indicating the mask area and the logical AND of D (X, Y) indicating the presence / absence of the luminance change detection are obtained, and the luminance change is obtained only in the monitoring target area. Pixel Extract only By setting the mask area as described above, an area in which the movement (movement) of an object other than the detection target is predicted to be detected is excluded from the monitoring target, and a false report is prevented from being issued.

Then, the error pixel counter 6 is in the monitoring target area,
Needless to say, pixels having a luminance change equal to or greater than the gradation difference j, that is, E
Number of pixels where (X, Y) = 1 And the number of pixels En counted up
The (error pixel) is compared with an alarm lower limit pixel number k (a threshold value of the number of pixels) arbitrarily set from outside, and an alarm is generated only when the counted pixel number En exceeds the alarm lower limit pixel number k. The signal AL is output to the alarm counter 7.

That is, when there are only a few pixels in the monitoring target area where the luminance change is greater than or equal to the gradation difference j, in other words, when the area of the screen where the gradation change is small, for example, For example, when the movement of a smaller object is detected, if the alarm signal AL is output as the detection of the desired moving object, an erroneous report is generated. The change in luminance is invalidated.

When the counted number of pixels En exceeds the alarm lower limit pixel number k and the alarm signal AL is output, the alarm counter 7 counts the alarm signal AL and sets the number of consecutive times 1 that can be arbitrarily set from the outside. The count value of the alarm signal AL is compared with the count value, and when the alarm signal AL is output continuously for a continuous number of times or more, an alarm (alarm) is issued to the outside as a moving object detection signal to notify the detection of the moving object.

As described above, an alarm is issued to the outside only when the output of the alarm signal AL continues for a predetermined number of times 1 or more, instead of issuing an alarm to the outside for each alarm signal AL. Alarm can be prevented from being output as a false alarm. That is, for example, when detecting the intrusion of an object into the field of view of the sensor 1, the number of continuations 1 is set to 1, and an alarm is issued to the outside each time the number of error pixels exceeds the alarm lower limit pixel number k. Similarly, when detecting an object moving in the field of view of the sensor 1, the number of error pixels is similarly set to the alarm lower limit pixel number k.
If an alarm is issued to the outside every time the sensor exceeds the limit, an alarm is output to the outside even when an object crosses the corner of the field of view of the sensor 1, which is a false alarm.

Therefore, only when the object to be detected moves within the detection field of view and the movement of the object is continuously detected, the output of the alarm signal to the outside is permitted,
For example, when the number of consecutive times 1 is set to 10, an alarm (moving object detection signal) is output to the outside only when the moving object detection based on the luminance difference output becomes 10 or more consecutive times. An alarm is output to the outside only when the movement of an object inside is detected, and no alarm is output to the outside even when it is detected that the object crosses the corner of the field of view.

If the number of continuous times 1 is set to 1 and an alarm is issued to the outside for each alarm signal AL, it is erroneously determined that a change in brightness due to lightning or lighting on / off is caused by a moving object,
Although there is a possibility that a false alarm is issued, it is possible to prevent an alarm from being output to the outside by a single phenomenon such as lightning or turning on / off the lighting by appropriately setting the number of continuous times l.

Incidentally, when the alarm signal AL is interrupted, the count number is reset to zero.

The capture of the latest image to be compared with the image stored in the video memory 4 can be variably controlled by a comparison interval i arbitrarily set from the outside.
When conforming to the NTSC system with Hz, for example, a frame image every 33 ms is set as the minimum unit, images are compared every 33 ms at the shortest, and the comparison time is set to be longer by a multiple of 33 ms. When the speed of the moving object is high, the comparison interval is shortened, and when the speed is slow, the comparison interval is lengthened.

The update timing of the stored image in the video memory 4 may be synchronized with the latest image capture period so that the previously captured image is always compared with the latest image. May be updated once every several times of image capture, so that the image in the video memory 4 may be compared with the latest image several times thereafter. You may make it keep.

Here, the control content of the moving object detection based on the luminance difference output as described above will be described with reference to the flowchart of FIG.

First, in S1, it is determined whether or not it is the comparison timing of the image determined by the comparison interval i, and if it is the comparison timing, the process proceeds to S2 to clear the error pixel counter 6.

Then, reading and updating of the reference image are performed in S3
In S4, it is determined whether or not the luminance difference for each pixel obtained by comparing the read reference image with the comparison image is equal to or greater than j gradation.

If the gradation difference is equal to or greater than j, the determination as to whether or not the pixel is a mask area is made in S5 based on, for example, whether or not 0 is set to the current pixel in the mask image. In some cases, the error pixel counter is incremented in S6.

On the other hand, when the tone difference is less than j or when the tone difference is included in the mask area, the process proceeds to S7 without incrementing the number of error pixels in S6.

In S7, it is determined whether or not the processing of S3 to S6 has been performed for all the pixels of the screen.If the processing of all the pixels has not been completed, the process returns to S3 and the processing is repeated.When the processing of all the pixels has been completed, Proceed to S8.

In S8, it is determined whether or not the number of error pixels counted in S6 exceeds k pixels, and when the number of error pixels exceeds k pixels (when the alarm signal AL is output in response to the determination). In step S9, go to alarm counter 7
(Continuous counter) is incremented, and in the next S10,
It is determined whether or not the count has exceeded a predetermined number l.

When it is determined in S10 that the count number has exceeded the predetermined number l, the process proceeds to S12 and outputs an alarm to the outside.
If the count is equal to or less than the predetermined number l, the process proceeds to S13, and no alarm is output to the outside.

If it is determined in step S8 that the number of error pixels is equal to or smaller than k pixels, the process proceeds to step S11 to clear the alarm counter, and then proceeds to step S13 to output no alarm to the outside.

Here, detection of a moving object by the above configuration is performed.
This will be briefly described again with reference to the drawings showing specific examples.

FIG. 3 shows a case where the brightness of each pixel converted by the A / D converter 3 is set to two gradations and the number of pixels is extremely smaller than a practical level for the sake of simplicity. Indicates bright and 0 indicates dark.

First, as a reference image which is an image before a predetermined time ΔT,
An image with a bright central part is stored.
The latest comparison image is an image in which the bright area is shifted to the left by one pixel with respect to the reference image. Therefore,
As shown in the next difference image (brightness difference output), the pixel in which the luminance change (gradation difference) has occurred is a pixel that has newly become 1 (brightened pixel) in the latest image and a reference image. In this case, the pixel which was 1 but became 0 (darkened pixel)
And the absolute value of the difference between the same coordinate pixels of the reference image and the comparison image is obtained as a pixel to which 1 is set.

On the other hand, as a mask map M setting indicating a monitoring target area, 1 is set and a pixel indicating a mask area is set in the lower half of the screen as shown in the figure, and 1 is set in the difference image. Among the pixels for which the generation of the gradation difference is specified and included in the lower half of the screen, those that are masked by this mask area, and as a result, those included in the upper half of the screen in the difference image Only the pixels that have changed in luminance remain. The pixel in which the remaining 1 is set corresponds to the error pixel, and the number of error pixels 2
Signal exceeds alarm lower limit pixel number k, alarm signal AL
Is output.

And the alarm signal AL output in this way
However, if the number of consecutive times becomes equal to or more than 1, the fact that a moving object has been detected is output to the outside for the first time. If the number of consecutive times 1 is set to 1, an alarm is output to the outside every time a moving object is detected.

Further, the state of the difference image and the occurrence of the alarm signal AL are
The case where the reference image is updated for each image comparison will be described with reference to FIG. 4. First, when the initial screen is the uninhabited room A and the intruder enters (B), the difference image between A and B is The alarm signal AL is transmitted because the luminance has changed due to the change in luminance by the entire intruder. Next, when the intruder moves indoors (C), also in this case, since the luminance changes due to the movement of the intruder between B before movement and A after movement, the alarm signal AL is also transmitted here. However, if the intruder stops moving at the moved position (D),
Since the intruder is located at the same position with respect to the previous time C and no luminance change occurs, the alarm signal AL is not transmitted. Also,
If a state change such as raising and lowering the arm is indicated even if it does not move (E), only the part of the arm indicates a change in luminance, and the alarm signal AL is also transmitted. Further, even when the intruder leaves the room, an alarm signal AL is transmitted because the brightness of the entire intruder changes depending on whether the intruder is in the room or not (F).

In the example shown in FIG. 4, the description is simplified by lengthening the comparison interval i. When the presence or absence of an intruder is detected, the comparison interval i is actually set to be shorter. ,
It is preferable that the image comparison is performed many times during the comparison of (B), the movement of the intruder is captured during that time, and the alarm signal AL is output each time. By the detection, for example, the alarm signal AL should be continuously output during the transition from the state (A) to the state (B), and the alarm signal should be output to the outside at least once.

FIG. 4 shows a case where the reference images are sequentially updated. However, as shown in FIG. 5, the reference images may be fixed.

In the example shown in FIG. 5, a state where a rectangular object is set in a room is imaged as a reference image. On the other hand, when a person enters the room (A), a luminance difference occurs due to the image of the person. Alarm signal AL is output, and
When such an intruder approaches the rectangular object (B), an intruder that is not in the reference image is also imaged, so that an alarm signal AL is output. When the intruder moves the rectangular object (C), a difference image is obtained due to the presence of the intruder and the movement of the rectangular object, and the alarm signal AL is also output. When the intruder exits, although the intruder does not exist, the difference image is output and the alarm signal AL is output because the position of the rectangular object moves with respect to the position in the reference image.

In FIG. 3, the number of pixels is set to about 20 for simplification of the description, but actually, a sensor having tens of thousands of pixels is used. However, experiments have shown that the number of pixels of about 15000 to 16000 is appropriate as the number of pixels that can balance the securing of detection accuracy and the avoidance of false alarms.

In FIG. 3, the mask region and the non-mask region are divided at the top and bottom of the screen. However, a plurality of regions can be set as the mask region or the non-mask region. It is also possible to designate not only a rectangle but also a region surrounded by an arbitrary curve with a pixel as a minimum unit, and it is also possible to designate a plurality of regions as a mask region and a non-mask region.

In setting the mask area (mask map), it is preferable that a monitor area or a non-monitoring area can be set by a cursor instruction in a state where a screen of the monitoring place is projected on an attached monitor. As shown in the figure, on the initial screen without the mask area designation, both ends of the straight line are designated by the cursor to draw a straight line, and the area surrounded by the straight line is displayed as a masked area on the screen as a solid area. As shown in the figure, the entire initial screen is designated as a mask area, and the entire screen is painted. In this state, a non-mask area (monitoring area) is designated by surrounding it with a straight line as described above, and the enclosed area is the monitoring area. By instructing that there is, the designated area is cleared from the filled state, and the mask area and the non-mask area are clearly displayed on the monitor There.

Further, once the detection conditions such as the gradation difference j, the mask area map M, the alarm lower limit pixel number k, and the number of consecutive times 1 are set,
Since the detection conditions are not changed frequently, the condition setting unit including the monitor and the detection unit that detects the moving object based on the set conditions can be divided into two parts. May be installed together with the sensor 1 (television camera) at a monitoring place. In this case, after installation, the condition setting cannot be changed unless a condition setting unit is prepared. This can prevent inadvertent change of the condition from impairing the reliability of the monitoring device, and can be achieved by replacing the condition setting unit with a plurality of detection units. To reduce costs by sharing
The size of the installed detection device can be reduced.

In the setting of the detection condition performed by connecting the condition setting unit, the detection condition variably set by a button operation or the like is displayed as a numerical value (a mask region map indicates a region overlapping the imaging screen) on an attached monitor. On the other hand, if the actual detection result (the number of error pixels, etc.) is also displayed as a numerical value on the monitor together with the screen of the monitoring place, the condition can be easily set according to the monitoring place and the monitoring situation.

Here, an example in which the detection unit and the condition determination unit are configured to be separable as described above will be described with reference to FIG. 8. The detection unit 11 includes an image signal from a video camera (two-dimensional imaging sensor 1) 12. The detection circuit 13 that receives this image signal processes the luminance difference output based on the processing conditions stored in the nonvolatile memory 14 and detects the moving object. . Detection unit 11 and condition setting unit
The setting data stored in the nonvolatile memory 14 by the switch device on the condition setting unit 15 side is connected to the shift register 17 of the detection unit 11, the composite cable 16, and the condition setting unit 15. Is transferred to the setting circuit 19 via the shift register 18.

The condition setting unit 15 includes the same detection circuit 20 as the detection unit 11, and performs the same detection operation based on the setting data transferred from the detection unit 11 and the image signal input via the plurality of cables 16. Do.

The setting data and the detection result data are synthesized into a video signal by a character generation circuit 21 and a character synthesis circuit 22, and displayed as a numerical value together with a monitoring image on a monitor 23 provided with a condition setting unit 15. The data transferred to the condition setting unit 15
The data can be changed by the operation switch 24, and the changed data is displayed on the monitor 24 in real time.

When the setting data (processing conditions) is finally set by the operation switch 24, the updated data is transmitted to the detection unit 11
The transfer is performed in a direction opposite to the forward path of the transfer from the controller 11, and is stored in the nonvolatile memory 14 of the detection unit 11.

When the updated setting data is stored in the non-volatile memory 14, the moving object can be detected by the detecting unit 11 alone. Therefore, by disconnecting the composite cable 16, the video camera 12 and the detecting unit can be detected.
When the detection device consisting of 11 functions alone and the setting data stored in the non-volatile memory 14 is to be updated, the condition setting unit 15 is connected via the composite cable 16.

Although the condition setting unit 15 is also provided with the detection circuit 20 in FIG. 8, the detection circuit is provided only on the detection unit 11 side, and the detection based on the setting data updated by the condition setting unit 15 is performed by the detection unit 11.
It is also possible to configure so that the result is transferred to the condition setting unit 15 side.

Further, in the above embodiment, one kind of detection condition such as the gradation difference j, the mask map M, the number k of alarm lower limit pixels, and the number of consecutive times 1 is set, and based on these, the current image and the time before the predetermined time ΔT are set. The comparison with the image was performed, but different detection conditions were set separately, and the processing of the luminance difference output based on the plurality of detection conditions (processing conditions) was performed in parallel. An alarm may be issued to the outside when the number of continuous alarm signals AL counted by the counter 7 exceeds the continuous number l.

Furthermore, a plurality of reference images having different update timings are provided,
The plurality of brightness difference outputs are obtained by comparing these reference images with the latest comparison image, and an alarm signal is generated based on each brightness difference output, and the alarm signal is counted by the alarm counter 7. You may comprise.

<Effects of the Invention> As described above, according to the present invention, a two-dimensional imaging unit for imaging a monitoring location is provided, and an image signal at the present time and an image signal at a time earlier by a predetermined time are taken by the two-dimensional imaging unit. To extract the luminance difference output,
In a monitoring system configured to detect a moving object in an image captured based on the luminance difference output, an alarm is issued to the outside only when moving object detection based on the luminance difference output is continued for a predetermined number of times or more. Because we configured
When detecting an object that moves in the monitoring field of view, a false alarm is issued when the object crosses the corner of the field of view, or a false change in brightness due to lightning or lighting on / off is detected as a moving object, etc. However, it is possible to avoid false reports due to a single event, and to improve the reliability of the monitoring system.

[Brief description of the drawings]

FIG. 1 is a system block diagram showing one embodiment of the present invention,
FIG. 2 is a flowchart showing the detection control contents of the system shown in FIG. 1, FIG. 3 is a diagram for explaining the detection processing contents in the above embodiment, and FIGS. 4 and 5 are detection processing in the above embodiment respectively. FIGS. 6 and 7 each show an example of mask area setting, and FIG. 8 is a block diagram showing an example in which a detection unit and a condition setting unit are configured to be separable. FIG. 1 2D imaging sensor 3 A / D converter 4 Video memory 5 Mask memory 6 Error pixel counter 7 Alarm counter j Tone difference (luminance Difference), M: mask area map, k: lower limit number of alarm pixels, l: number of consecutive times, i: comparison interval

Continuation of the front page (58) Field surveyed (Int.Cl. ⁶ , DB name) G01V 9/04 G01P 13/00

Claims (1)

(57) [Claims] 1. A two-dimensional image pickup means for picking up an image of a monitoring location, wherein a current difference image signal picked up by the two-dimensional image pickup means is compared with an image signal at a point in time preceding by a predetermined time to extract a luminance difference output. A monitoring system configured to detect a moving object in an image captured based on the luminance difference output, wherein an alarm is issued to the outside only when moving object detection based on the luminance difference output is continued for a predetermined number of times or more. A monitoring system characterized in that it is configured as described above.