JPH03138595A - Moving body detecting device - Google Patents

Abstract

PURPOSE:To evade misinformation due to a single phenomenon and to improve the reliability by sending out a detection signal only when the frequency of continuous moving body detection based upon a brightness difference output exceeds a specific value. CONSTITUTION:The latest digital image signal B is compared with an image A which is stored 4 and a specific time before in picture element units to calculate the gradation difference C (brightness difference output), which is compared with the gradation difference (j) which is the threshold value of the brightness difference to remove picture elements which are less than the gradation difference (j) as brightness variation which is irrelevant to a moving body. Then brightness variation detection outside an area to be monitored is made ineffective by using an optionally set mask area map M and the misinformation is evaded. Then the number En of picture elements which vary in brightness by more than the gradation difference (j) in the area to be monitored is counted 6 and an alarm signal AL is outputted only when the counted number of picture elements exceeds the alarm lower-limit number (k) of picture elements which is its threshold value; and an alarm counter 7 counts the signal AL and sends an alarm for a moving body detection signal when the signal AL is outputted continu ously by more than the number l of continuous times, thereby indicating the detection.

Description

DETAILED DESCRIPTION OF THE INVENTION (Industrial Application Field) The present invention relates to a moving object detection device, and more specifically, the present invention relates to a moving object detection device. The present invention relates to a moving object detection device.

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

By detecting moving objects in a wide three-dimensional area, which is difficult to do with such -dimensional sensors, this sensor is used in surveillance systems that can provide security and surveillance over a wide area by bringing the situation closer to what is being monitored with the human eye. As a method for identifying moving objects from captured images using a two-dimensional image sensor such as a television camera, a method is being considered.

However, this type of moving object detection requires either a manned system in which a security guard constantly monitors images captured by a television camera to detect moving objects, or a system in which such images are temporarily recorded with a video table recorder. , which was a system that later discovered moving objects from reproduced images. here,
The image may be a moving image, or it may be a still image taken at predetermined time intervals for economic reasons, but in either case, the human eye will ultimately determine the presence or absence of a moving object. It is supposed to be.

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 monitor, especially when monitoring the monitors of cameras installed in multiple locations at the same time. There was also the risk of missing the presence of moving objects such as intruders.

A moving object detection device that solves this problem compares the current image signal captured by a two-dimensional image sensor with the image signal a predetermined time ago, and moves based on the occurrence of a brightness difference between the two image signals. A method for automatically detecting an object has previously been proposed (see Japanese Patent Laid-Open No. 64-65486, etc.).

In other words, if there is no moving object in the captured image, the images at the current moment and the image a predetermined time ago should be approximately the same, and no large brightness changes will occur. When a person enters a space, a large brightness difference occurs between the image data captured before and after the person enters the space, so the presence or absence of a moving object is detected by capturing the occurrence of this brightness difference.

<Problems to be Solved by the Invention> Incidentally, in the unmanned monitoring system using the two-dimensional image sensor as described above, an alarm is issued to the outside each time a moving object is detected based on the brightness difference output. When detecting the intrusion of an object into the monitoring field of view, an alarm should be output to the outside each time a moving object is detected based on the brightness difference output as described above. When detecting objects, even if an object crosses the corner of the field of view, an alarm may be output, resulting in a false alarm.

In other words, when you want to detect an object moving within the monitoring field of view, if you issue an alarm every time a detection is based on the luminance difference output, it is possible that the object is moving within the field of view or has crossed the edge of the field of view. It was not possible to determine whether an object was detected or not, and even if various processing conditions (detection sensitivity, etc.) for the brightness difference output were set, the determination could not be made with high accuracy.

In addition, if the configuration is such that an alarm is issued to the outside each time a moving object is detected based on the luminance difference output, even when detecting an object entering the monitoring field of view, lightning or lighting may be turned on.・If the surrounding brightness changed in a short time when the device was turned off, there was a risk that this would be interpreted as a change in brightness caused by a moving object and a false alarm would be output.

The present invention was created in view of the above problems, and prevents false alarms in the event of a single event by improving the characteristics of the alarm that is output to the outside every time a moving object is detected based on the brightness difference output. The purpose is to improve the reliability of the detection device.

<Means for Solving the Problems> Therefore, in the present invention, a two-dimensional imaging means is provided, and the image signal at the present moment taken by the two-dimensional imaging means is compared with the image signal at a time a predetermined time before, and the luminance is determined. In a moving object detection device that extracts a difference output and detects a moving object in a captured image based on the brightness difference output, when the number of consecutive moving object detections based on the brightness difference output reaches a predetermined number or more. The configuration was configured to allow output of the moving object detection signal only to the outside.

<Operation> According to this configuration, the current image signal captured by the two-dimensional imaging means and the image signal at a predetermined time ago are compared, and a luminance difference output is extracted. Here, the moving object in the captured image is detected based on the brightness difference output, but only when the number of consecutive detections of the moving object based on the comparison of the two images is a predetermined number of times or more, the moving object is detected. Since the output of the object detection signal is permitted, when the number of consecutive moving object detections is less than the predetermined number of times, the detection signal is not output to the outside even if a moving object is detected, and a single moving object is detected. In this case, no detection signal is output to the outside.

<Example> The present invention will be explained in detail below.

FIG. 1, which shows one embodiment, shows a monitoring system that warns the outside of the presence of a moving object at a specific monitoring location. The dimensional imaging sensor 1 is installed facing the specific monitoring location, and photographs the monitoring location and outputs an analog image signal.

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

The latest (current) digital image signal B is compared pixel by pixel with the image A stored in the video memory 4 at a predetermined time ΔT before, and the luminance gradation difference C(
luminance difference output) is calculated. Here, if a pixel is represented by X and Y coordinates, the gradation difference (brightness difference) C is
(X, Y)=A (X, Y)-B (X, Y)l.

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

Once the pixel-by-pixel gradation difference C (X, Y) between the latest image (comparison image) and the image at a predetermined time period ΔT (reference image) stored in the video memory 4 is calculated, a The gradation difference j, which is the threshold value of the brightness difference to be set, and the gradation difference c (x, y) are compared for each pixel, and the gradation difference j
Select only the pixels whose brightness has changed by the above amount, and classify all pixels into pixels whose brightness has changed (1 is set as data) and pixels whose brightness has not changed (O is set as data). Let it be binarized data D (X, Y). Therefore, even if the brightness changes during a predetermined period of time ΔT, pixels for which the change is less than 1 gradation difference will be treated as if there was no change in brightness. This is to prevent erroneous detection of brightness changes unrelated to the moving object, and the sensitivity of the moving object can be adjusted using the gradation difference j.

After all pixels are binarized based on the presence or absence of a luminance change within a predetermined time ΔT as described above, next, the mask area (mask area map) stored in the mask memory 5, which is also arbitrarily set from the outside, is M) disables brightness change detection in areas that are not monitored. Specifically, for example, a pixel with a brightness change is set to 1, a pixel with no gradation change is set to 0, a mask pixel that is not to be monitored is set to 1, and a non-mask pixel that is to be monitored is set to 0. If M (X.

Y) and D(x,
y), the pixel E (X, Y) = D (X, Y) whose brightness changed only in the area to be monitored
・Extract only V1 TYY. By setting the mask area as described above, an area in which movement of an object other than the detection target is expected to be detected is excluded from the monitoring target, thereby avoiding issuing a false alarm.

Then, the error pixel counter 6 detects a pixel in the monitoring target area that has a luminance change of more than the gradation difference j, that is, E
(X, Y) -1 pixel number En = n = runt-up pixel number En (error pixel) is compared with alarm lower limit pixel number k (pixel number threshold) arbitrarily set from the outside. , the alarm signal AL is sent only when the counted up pixel number En exceeds the alarm lower limit pixel number k,
Output to alarm counter 7.

That is, when there are few pixels in the monitoring target area that have a luminance change of more than the gradation difference j, in other words, when the area on the screen where the gradation change has occurred is small, for example, if there is a moving object to be detected. If the movement of an object smaller than the object is detected, outputting an alarm signal AL as a detection of a desired moving object will result in a false alarm. This is invalidated for brightness changes.

When the counted-up pixel number En exceeds the alarm lower limit pixel number k and an alarm signal AL is output, the alarm counter 7 counts this alarm signal AL, and sets a continuous count that can be arbitrarily set from the outside. Number of times! and the count value of the alarm signal AL, and when the alarm signal AL is output for two or more consecutive times, an alarm is issued as a moving object detection signal to the outside to notify the detection of a moving object.

In this way, instead of issuing an external alarm for each alarm signal AL, the alarm signal AL is output a predetermined number of times! By issuing an alarm to the outside only when the above occurrences occur continuously, it is possible to prevent an alarm from being output to the outside due to a single phenomenon, resulting in a false alarm. That is, for example, when detecting the intrusion of an object into the field of view of the sensor 1, it is sufficient to set the consecutive number l to 1 and issue an alarm to the outside every time the number of error pixels exceeds the alarm lower limit pixel number k. However, when detecting an object moving within the field of view of sensor 1, the number of error pixels is equal to the alarm lower limit pixel number k.
If the configuration is such that an alarm 1 is emitted to the outside each time the object exceeds the threshold, an alarm will be output to the outside even if an object crosses the corner of the monitoring field of the sensor 1, resulting in a false alarm.

For this reason, the device is configured so that output of an alarm signal to the outside is permitted only when the object to be detected moves within the detection field of view and movement of the object is detected continuously. When ℃ is set to 10, an alarm (moving object detection signal) is output to the outside only when a moving object is detected 10 times or more in a row based on the luminance difference output, and an object within the monitoring field of view is detected. An alarm is output to the outside only when movement of the object is detected, and an alarm is not output to the outside even if it is detected that the object has crossed the field of view.

In addition, if the configuration is such that the consecutive number of times 2 is set to 1 and an alarm is issued to the outside for each alarm signal AL, changes in brightness due to lightning or turning on and off lights will be incorrectly determined to be caused by a moving object.
Although there is a possibility of false alarms being issued, by appropriately setting the continuous number of times, it is possible to avoid outputting alarms to the external device 2 due to one-off phenomena such as lightning or turning lights on and off.

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

Furthermore, the latest image to be compared with the stored image on the video memory 4 is captured at a comparison interval i that is arbitrarily set from the outside.
For example, the field frequency can be variably controlled by
■If the NTSC system is compliant with 2, for example, 3
The minimum unit is a frame image every 3m5, and the shortest is 33
The image is compared every m5, and the comparison time is set to be longer by a multiple of 33m5, and when the speed of the object moving in the screen is fast, the comparison interval is shortened, and when it is slow, the comparison interval is lengthened.

Note that the update timing of the stored image in the video memory 4 may be synchronized with the capture cycle of the latest image, and the previously captured image and the latest image may be constantly compared. The image on the video memory 4 may be updated once every several times of image capture, and the image on the video memory 4 may be compared with the latest image several times thereafter. You may also leave it as is.

Here, the content of control for moving object detection based on the luminance difference output as described above will be explained based on the flowchart of FIG. 2.

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

Then, the reference image is read and updated pixel by pixel.
In S4, the brightness difference for each pixel, which is found by comparing the read reference image and the comparison image, is j
Determine whether there is a basic tone.

If the gradation difference is 1 or more, in s5 it is determined whether the area is a mask area or not, for example, by whether or not O is set for the current pixel in the mask image. At some point, an error pixel counter is incremented in S6.

On the other hand, when the gradation difference is less than j or included in the mask area 3 4 , the process proceeds to 37 without incrementing the number of error pixels in S6.

In S7, it is determined whether or not the processing of 83 to S6 has been performed for all pixels on the screen. If the processing of all pixels has not been completed, the process returns to S3 and the processing is repeated. When the processing of all pixels has been completed, Proceed to S8.

In S8, it is determined whether or not the number of error pixels counted in S6 exceeds the number of pixels, and if the number of error pixels exceeds the number of pixels (in response to such a determination, an alarm signal
is output), proceed to S9, increment alarm counter 7 (continuous counter), and start the next SIO.
Then, it is determined whether the count number exceeds a predetermined number l.

When it is determined in step 310 that the count number exceeds the predetermined number of times l, the process advances to S12 and an alarm is output to the outside, but the count number has exceeded the predetermined number of times! If it is below, S13
Proceed to and do not output alarms to the outside.

If it is determined in S8 that the number of error pixels is less than or equal to 5, the process proceeds to 311 to clear the alarm counter, and then proceeds to S13 to not output an alarm to the outside.

Here, detection of a moving object using the above configuration will be briefly explained again with reference to figures showing specific examples.

In FIG. 3, in order to simplify the explanation, the brightness of each pixel converted by the A/D converter 3 is set to two gradations, and the number of pixels is also extremely smaller than the practical level. Assume that 0 indicates brightness and 0 indicates darkness.

First, an image with a bright central part is stored as a reference image, which is an image taken a predetermined time ΔT before.On the other hand, in the latest comparison image, the bright area is shifted by one pixel to the left with respect to the reference image. This is an image of what is happening. Therefore, as shown in the following difference image (brightness difference output), pixels in which a brightness change (gradation difference) has occurred include pixels that have newly become 1 in the latest image (pixels that have become brighter); A pixel that was 1 in the standard image but becomes O (a darkened pixel), and a pixel that is set to 1 by taking the absolute value of the difference between the same coordinates 6 pixels in the standard image and comparison image. It is required as.

On the other hand, as the mask map M setting that indicates the area to be monitored, pixels are set to 1 to indicate that the area is a mask area, as shown in the figure, and 1 is set in the difference image. Among the pixels for which gradation differences have been identified, those included in the lower half of the screen are masked by this mask area, and as a result, only those included in the upper half of the screen are included in the difference image. remains as a pixel in which a luminance change has occurred. Then, the remaining pixels set to 1 correspond to error pixels, and if the number of error pixels 2 exceeds the alarm lower limit pixel number k, an alarm signal AL is sent.
will be output.

When the alarm signal AL output in this way reaches 1 or more consecutive times, it outputs for the first time that a moving object has been detected to the outside. In addition, the number of consecutive times mentioned above! If is set to 1, an alarm will be output to the outside every time a moving object is detected.

In addition, the state of the differential image and the occurrence of the alarm signal AL are
The case where the reference image is updated for each image comparison will be explained based on FIG. 4. First, when the initial screen is an unoccupied room A and an intruder enters the room (B), the difference image between A and B is , the change in brightness caused by the entire intruder is taken as a change in brightness, and an alarm signal AL is sent because there is a change in brightness. Next, when the intruder moves inside the room (C), in this case as well, the alarm signal AL is transmitted because the brightness changes between B before movement and A after movement due to the movement of the intruder.

However, if the intruder stops moving at the moved position (D
), the intruder is located at the same position as the previous time C and no change in brightness occurs, so the alarm signal AL is not transmitted.

Furthermore, if the arm shows a change in status such as moving up and down even if it does not move (E), only the part of the arm will show a change in brightness, and this will also cause the alarm signal AL to be transmitted. Further, even when the intruder leaves the room, the alarm signal AL is generated because the brightness of the intruder's entire body changes depending on whether he is in the room or not (F).

Note that in the example shown in FIG. 4, the comparison interval i is lengthened to simplify the explanation; when detecting the presence or absence of an intruder, the comparison interval i is actually set shorter, and for example, image (A) ,(
During the comparison in B), it is preferable that the image be compared many times, and the movement of the intruder during that time is captured and an alarm signal AL is output each time. For example, from state (A) to (
It should be set so that the alarm signal AL is continuously output during the transition to state B) and the alarm signal is output to the outside at least once.

Although FIG. 4 shows the case where the reference image is updated sequentially, the reference image may be fixed as shown in FIG. 5.

In the image shown in FIG. 5, a rectangular object is imaged in a room as a reference image, and when a person enters the room (A), a brightness difference occurs due to the image of the person. Alarm signal AL is output, and
When such an intruder approaches the nine rectangular object (B), an intruder who is not in the reference image is also imaged, so an alarm signal AL is output. Also, if the intruder moves a rectangular object (C)
, a differential image is obtained due to the presence of an intruder and the movement of a rectangular object, and an alarm signal AL is also output. When the intruder leaves, although the intruder does not exist, the position of the rectangular object has moved relative to the position in the reference image, so a difference image is output and an alarm signal AL is output.

In FIG. 3, the number of pixels is about 20 to simplify the explanation, but in reality, sensors for enemy pixels are used. However, the number of pixels that can balance between ensuring detection accuracy and avoiding false alarms is 15,000 to 1.
Experiments have shown that a pixel count of about 6000 is appropriate.

In addition, in Figure 3 above, the masked area and non-masked area are divided into the upper and lower parts of the screen, but it is also possible to set multiple areas as masked areas or non-masked areas, and the masked area and non-masked area are It is possible to designate not only a quadrangle but also a region surrounded by an arbitrary curve using pixels as the minimum unit, and furthermore, it is also possible to designate a plurality of regions as masked regions and non-masked regions.

When setting the mask area (mask map), it is preferable to configure the system so that the monitored area or non-monitored area can be set by pointing the cursor while displaying the screen of the monitored location on the attached monitor. For example, as shown in FIG. As shown in Figure 7, on the initial screen where no mask area is specified, a straight line is drawn by pointing both ends of the line with the cursor, and the area surrounded by this straight line is filled in and displayed on the screen as a mask area. As shown in the figure, the entire initial screen is specified as a mask area to fill the entire screen, and in this state, the non-masked area (monitoring area) is specified by surrounding it with a straight line as described above, and the enclosed area is the monitoring area. By instructing,
The specified area is cleared from being filled, and
The configuration may be such that the masked area and the non-masked area are clearly displayed on the monitor.

Furthermore, the gradation difference j, the mask M area map M, the alarm lower limit pixel number, and the number of consecutive times! Once the detection conditions are set, the detection conditions will not be changed frequently, so the condition setting section including the monitor and the detection section that detects moving objects based on the set conditions can be divided into two parts. After setting the detection conditions, set only the detection part to sensor 1 (TV camera).
It may also be installed at a monitoring location. In this case, after installation, the condition settings cannot be changed unless a condition setting section is prepared. This prevents damage to the reliability of the monitoring device due to careless changes in conditions, and also allows the condition setting section to be connected to multiple detection sections. It is possible to reduce costs by sharing the detection device with other devices, and also to reduce the size of the installed detection device.

When setting detection conditions by connecting the condition setting section,
The configuration is configured so that the detection conditions that are variably set by button operations etc. are displayed as numerical values on the attached monitor (the mask area map indicates the area that overlaps the imaging screen), while the actual detection results (number of error pixels, etc.) are also displayed. If the value is displayed as a numerical value 1 2 on the monitor together with the monitoring location screen, conditions can be easily set according to the monitoring location and monitoring situation.

Here, an example in which the detection section and the condition determination section are separable as described above will be explained with reference to FIG. 8. The detection circuit 13 that receives this image signal is configured to process the brightness difference output based on processing conditions stored in the nonvolatile memory 14 to detect a moving object. . The detection section 11 and the condition setting section 15 are connected by a composite cable 16, and the setting data stored in the nonvolatile memory 14 is transferred to the shift register 17. of the detection section 11 by a switch device on the condition setting section 15 side. Composite cable 169 condition setting section 15
The data is transferred to the setting circuit 19 via the shift register 18.

The condition setting section 15 includes the same detection circuit 20 as the detection section 11, and uses the setting data transferred from the detection section 11,
The same detection operation is performed based on the image signals inputted via the plurality of cables 16.

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

When the setting data (processing conditions) are finally set using the operation switch 24, the updated data is transmitted to the detection unit 1.
The data is transferred in the opposite direction to the transfer route from 1 to the detection unit 1.
1 nonvolatile memory 14.

When the updated setting data is stored in the nonvolatile memo, the detection unit 11 can detect a moving object by itself, so by disconnecting the composite cable 16, the detection device consisting of the video camera 12 and the detection unit 11 can be detected. When it is desired to update the setting data stored in the non-volatile memory 14, the condition setting section 15 is connected to the composite cable 16.
Connect via.

In FIG. 8, the condition setting section 15 is also provided with the detection circuit 20, but only four detection circuits are provided on the detection section 11 side, and the condition setting section 15 performs
Detection is performed on the detection unit 11 side, and the result is sent to the condition setting unit 1.
The configuration may be such that the data is transferred to the 5th side.

In addition, in the above embodiment, the tone difference j, the mask mask M,
One type of detection condition such as consecutive number of times 2 was set for each alarm lower limit pixel number, and based on these, the current image and the image at the time before a predetermined time ΔT were compared, but different detection conditions were set separately. and multiple detection conditions (processing conditions)
The processing of the luminance difference output based on the results is performed in parallel, and the resulting alarm signal A
When the number of consecutive alarm signals AL exceeds the number i of consecutive alarm signals, an alarm may be issued to the outside.

Further, it includes a plurality of reference images with different update timings, obtains the plurality of brightness difference outputs by comparing these reference images with the latest comparison image, and generates an alarm signal based on each brightness difference output, The configuration may be such that this alarm signal is counted by the alarm counter 7.

<Effects of the Invention> As explained above, according to the present invention, in a device that detects a moving object based on a luminance difference output obtained by comparing an image signal at a time before a predetermined time with an image signal at the present time, The configuration allows the output of a moving object detection signal to the outside only when the number of consecutive moving object detections exceeds a predetermined number, so when detecting an object moving within the monitoring field of view, it is possible to It is possible to avoid false alarms caused by one-off phenomena, such as false alarms being issued when an object crosses the street, or false detections of brightness changes due to lightning or lighting turning on and off as moving objects, and improve the reliability of the detection device. It can improve convertibility.

[Brief explanation of the drawing]

FIG. 1 is a system block diagram showing an 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 same embodiment as above, and Figs. 4 and 5 respectively show the detection processing in the above embodiment. 6 and 7 are diagrams each showing an example of mask area setting,
FIG. 8 is a block diagram showing an example in which the detection section and the condition setting section are separable. 1... Two-dimensional image sensor 3... A/D converter 4
...Video memory 5...Mask memory 6...
Error pixel counter 7...Alarm counter
j...Gradation difference (brightness difference) M...Mask area map k...Alarm lower limit number of pixels! ...Continuous number i...Comparison interval

Claims (1)

[Scope of Claims] A two-dimensional imaging means is provided, and a luminance difference output is extracted by comparing an image signal at the current time captured by the two-dimensional imaging means with an image signal at a time a predetermined time before, and the luminance difference In a moving object detection device that detects a moving object in a captured image based on an output, a moving object detection signal is sent to the outside only when the number of consecutive moving object detections based on the luminance difference output reaches a predetermined number or more. A moving object detection device characterized in that it is configured to permit output of.