JP3222392B2 - Anomaly detection method of monitoring target place using captured image - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a security system for detecting a change occurring in a monitoring target by using an image of a monitoring camera photographing a monitoring target location.

[0002]

2. Description of the Related Art FIG.
Shown in The monitoring device 100 is a monitoring device installed at a monitoring target location, and the monitoring device 120 is a monitoring device installed at a location where a monitor is present. Monitoring device 10
A communication line 109 is provided between the monitoring device 120 and the monitoring device 120. The communication line 109 transmits a change in an image generated in the monitoring target detected by the monitoring device 100 to the monitoring device 120, notifies a monitoring person, and performs image capturing. The change in the monitored object is displayed on the display device.

The monitoring apparatus 100 includes a CPU 101 for controlling the monitoring apparatus, a ROM 102 for storing a control program for the monitoring apparatus including image processing, a RAM 103 as a memory required for each processing, and an arithmetic processor 10 for high-speed operation.
4. Passive sensor 105 for detecting an abnormality in a monitoring target location by infrared rays, frame memory 106 for storing image data, monitoring (CCD) camera 107 for capturing an image of the monitoring target location and outputting image data, image data And a communication control unit 108 for transmitting an alarm signal. In conjunction with the detection of an abnormality by the passive sensor 105, the image captured by the monitoring camera 107 is compressed and transmitted from the communication control unit 108, and the communication control of the monitoring device 120 is performed via the communication line 109. Input to the unit 118.

The CPU 110 constituting the monitoring device 120,
The compressed data received is expanded and decompressed by the ROM 112, the RAM 113, and the arithmetic processor 114.
The information is displayed on the display device 117 via the frame memory 116, and the abnormality is notified to the observer by the operation of the alarm device 115.

[0005]

The guard area of the passive sensor 105 of the monitoring device 100 installed at the monitoring target location needs to be changed depending on the purpose of use. The image input range by the monitoring camera 107 is set in cooperation with the guard area of the passive sensor 105,
It is difficult to make a perfect match. For this reason, even if the passive sensor 105 detects a change in the alert area and inputs an image from the monitoring camera 107 that operates in conjunction with the passive area, the cause of the change caused by the passive sensor 105 reacting to the image is as follows. Sometimes it was not reflected.

[0006]

SUMMARY OF THE INVENTION The present invention has been made to solve the above-mentioned drawbacks of the prior art, and continuously captures images by a monitoring camera without performing detection by a passive sensor. The difference image between the previous image and the current image is calculated using the still image at the monitoring target location, and the difference image is compressed, and the compressed data amount exceeds a preset threshold. In order to detect a change that has occurred in the monitoring target location where the shooting is being performed.

[0007]

DESCRIPTION OF THE PREFERRED EMBODIMENTS Embodiments of the present invention will be described with reference to the drawings. FIG. 1 is a block diagram showing a configuration of a monitoring device according to the present invention. The first monitoring device 1 has a CPU
1, a ROM 2, a RAM 3, an arithmetic processor 4, a frame memory 6, a surveillance (CCD) camera 7, and a communication control unit 8, except that the surveillance device 100 according to the prior art does not include the passive sensor 105. And the same configuration. In addition, the second monitoring device 2
PU 11, ROM 12, RAM 13, arithmetic processor 1
4, alarm device 15, frame memory 16, display device 1
7. The communication control unit 18 has the same configuration as the monitoring device 120 according to the related art, and has the same various functions.

The passive sensor 105 of the monitoring device 100
Instead, software having a function of detecting an abnormality is stored in the ROM 2 of the first monitoring device 1, and details thereof are as follows. FIG. 2 shows a process of inputting a plurality of still images of a monitoring target location, obtaining differential image data thereof, calculating upper and lower threshold values from an average value of compressed data lengths subjected to compression processing, and setting initial values. It is a flowchart.

In FIG. 2, a still image of a monitoring target location is input from the monitoring camera 7 to the frame memory 6 (step S1), and it is determined whether or not the obtained image is the first image (step S2). . If it is the first image, the image data on the frame memory 6 is transferred to the delay memory (R
(A memory temporarily set in AM3) (Step S)
3) Then, the process returns to step S1. If it is not the first image data, a difference image between the image data on the frame memory 6 and the image data on the delay memory is obtained (step S4), and the image data on the frame memory 6 is stored in the delay memory. (Step S5). The difference image data is subjected to compression processing (data compression by performing frequency conversion, quantization, and variable length coding processing) to determine a compressed data length (step S6), and determines whether the n-th compressed data length has been determined. It is determined (step S7). When the compressed data length for n sheets is obtained, the average value is calculated (step S8), and the upper limit threshold value and the lower limit threshold value are calculated from the average value (step S9). If the compressed data length of the n-th sheet has not been obtained, the process returns from step S7 to step S1, and the above-described steps are repeatedly executed to obtain the upper and lower limits of the threshold (initial value).

A threshold value calculated in the step shown in FIG. 2 is set in advance, a difference image is obtained from a still image of the target location to be monitored, and the compressed data length is compared with the threshold value to determine the abnormality of the target location. FIG. 3 is a flowchart showing the detection process. In FIG. 3, the monitoring camera 7
, The still image of the monitoring target location is input to the frame memory 6 (step S11), and difference image data between the image data on the frame memory 6 and the image data on the delay memory is obtained (step S12). The image data on the frame memory 6 is stored in the delay memory (step S13). The difference image data is compressed to obtain a compressed data length (step S14), and it is determined whether the compressed data length exceeds an upper limit or a lower limit of a threshold (step S15). Determines that there is a change in the image and performs an alarm process (step S16). If not exceeded, the average value of the compressed data length of n sheets including the current compressed data length is calculated (step S17), and the upper and lower thresholds are calculated from the average (step S1).
8). The threshold value is updated in step S18, and the above-described steps are repeated from step S11 again to monitor whether or not the monitoring target location is abnormal.

As described above, the first monitoring device 1 is provided with software for detecting the presence / absence of an abnormality in the monitoring target place. When an abnormality is detected, a signal from the first monitoring device 1 is output. As a result, the alarm device 15 of the second monitoring device 2 is activated. In conjunction with this, the compressed image data from the first monitoring device 1 is expanded and restored in the second monitoring device 2 and displayed on the display device 17. Will be displayed.

FIG. 4 is a graph showing the relationship between the amount of compressed data and the temporal change in the upper and lower limits of the threshold.

[0013]

As described above, the method of detecting an abnormality in a monitoring target location using a captured image according to the present invention is applied to a still image of a monitoring target location captured by a monitoring camera and stored in a frame memory. Detects a difference between the previous image and the current image, and compares the amount of compressed data obtained by compressing the difference image with a preset upper or lower threshold to detect an abnormality in the monitoring target location. And display the still image. Therefore, since a defect due to a difference between the guard area of the passive sensor and the image input range of the monitoring camera can be completely removed, it is possible to accurately detect the abnormality occurring in the monitoring target location.

[Brief description of the drawings]

FIG. 1 is a block diagram showing a configuration of a monitoring device according to the present invention.

FIG. 2 is a flowchart showing a process of setting an upper limit and a lower limit of a threshold.

FIG. 3 is a flowchart illustrating a process of detecting an abnormality in a monitoring target location.

FIG. 4 is a graph showing the relationship between the amount of compressed data and the threshold over time.

FIG. 5 is a block diagram showing a configuration of a monitoring device according to a conventional technique.

[Explanation of symbols]

 1,11 CPU 2,12 ROM 3,13 RAM 4,14 Arithmetic processor 6,16 Frame memory 7 Surveillance camera 8,18 Communication control unit 9 Communication line 15 Alarm device 17 Display device

Claims (1)

(57) [Claims] 1. A still image of a monitoring target place which is continuously photographed by a monitoring camera of a first monitoring device installed in a monitoring target place and stored in a frame memory is compared with a still image of one previous picture. A difference image from the current still image is calculated, and the amount of compressed data obtained by compressing the difference image is compared with a preset threshold to detect an abnormality that has occurred in the monitoring target location, and to determine the location where the monitor is located. A monitoring object using a photographed image, characterized in that an abnormal signal is transmitted to a second monitoring device installed in the apparatus and a warning is issued, and a still image indicating the abnormality is displayed on the second monitoring device. Location anomaly detection method.