JP2001224011A - Abnormality supervisory method and system - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide an abnormality detection method and system that can relieve the load on a supervisor with high precision in order to solve problems of conventional systems in many of which among various ways of image supervision the supervisor always supervises the entire image and discriminates whether or not abnormality is in the image thereby making less improvement of accuracy of the supervisory. SOLUTION: In this method, an area surrounded on an image of a supervisory object by line segments is set as a supervisory notice area, the supervisor takes notice of the line segments set as the supervisory area for the abnormity of the supervisory area, supervises a state change in the line segments for a predetermined period and discriminates whether or not the supervisory notice area is abnormal based on the state change of the line segments. Furthermore, time series image data for a predetermined period before and after the occurrence of abnormity are stored.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to an abnormality monitoring method for monitoring an object to be monitored in an art museum, a bank, or the like using a photographing device such as a TV camera, and automatically detecting an abnormality of the object to be monitored. Related to the device.

[0002]

2. Description of the Related Art Conventionally, the most frequently used anomaly monitoring (or anomaly detecting device) of a remote place is a TV camera installed to a monitoring target area, a TV monitor for displaying a monitoring image, and various components and devices. A device configured by a control device that controls an image signal is used. In many cases, an observer looks at an image photographed on a TV monitor, directly monitors for the occurrence of an abnormality, and makes an abnormality determination.

Further, there is a method in which an advanced remote location abnormality monitoring (abnormality detection device) automatically recognizes the occurrence of an image change in a monitoring target area and transmits an image when the change occurs to a monitor.

[0004] As such a remote place abnormality monitoring method,
(1) Japanese Patent Application Laid-Open No. 5-64199, "Image Monitoring Apparatus",
(2) Japanese Patent Application Laid-Open No. Hei 8-221577, "Moving Object Detection / Extraction Apparatus and Method", and (3) Japanese Patent Application Laid-Open No. 8-134506, "Moving Object Combination Detection Apparatus and Method". In particular, the above-mentioned (1) Japanese Patent Application Laid-Open No. 5-64199 discloses a technique of monitoring a change in an image of a monitoring target area, dissimilarity information indicating the degree of image change in the monitoring target area, and an identifier for identifying the monitoring target area at that time. , And an image change detection unit that generates and outputs the same.

[0005]

However, the most frequently used conventional monitoring technique is a method in which a monitoring person judges an abnormality detection by looking at a monitor. In this case, the observer cannot continuously concentrate on the monitoring target alone because the entire monitoring image is to be monitored, and maintains the accuracy of abnormality detection due to a human error such as a decrease in concentration. It is difficult.

Further, in the case of a monitoring device that automatically recognizes the occurrence of an image change in a monitoring target area and transmits the image in which the image change has occurred to a monitoring person, the detection accuracy is high, but the focus is on the monitoring target. Monitoring is difficult. Also,
Anomalies due to unnecessary image changes may be detected. Further, since many systems do not have a device for storing the detected image data as a still image, the image data cannot be used for recording an event or investigating the cause of an abnormal phenomenon.

An object of the present invention is to set a monitoring area by surrounding a plurality of straight lines or curves with respect to an image of an object to be monitored.
The detection of an abnormality in a monitoring target focuses on a line segment that sets the monitoring area, and provides a monitoring target abnormality monitoring method and apparatus in which the line segment abnormality is determined to be a monitoring target abnormality. is there.

[0008]

The present invention is characterized by having the following three functions in order to solve the above-mentioned problems. The first function is that a monitoring target is surrounded by an arbitrary line segment in a monitoring image to be a monitoring target area, and abnormality of the monitoring target is detected by a change in data of the line segment. The second function is to have a function of storing image data at the time of occurrence of an abnormality in which line segment data has changed as a still image. The third function is to accumulate digital image data for a predetermined period (time) before and after the occurrence of an abnormality.

More specifically, a monitoring area of interest is set around a monitoring target object by a plurality of line segments in a captured monitoring target image, and when a change in the state of the line segment is detected, the monitoring target object is detected. It is determined that the image data is abnormal, and image data for a predetermined time before and after the occurrence of the line segment abnormality is stored as still image data.

[0010]

Embodiments of the present invention will be described below. FIG. 1 shows a configuration diagram of an abnormality monitoring device that performs monitoring from a remote place as one embodiment of the present invention. The remote monitoring unit 1 includes a photographing device 2 for photographing a monitoring target.
And the terminal device 3. The terminal device 3 is an image change detection device 4.
And a communication device 9. First, the remote place monitoring unit 1 will be described.

An image photographed by the photographing device 2 is analog /
An image change is detected and detected via an image input device 5 for digital conversion (A / D conversion) and taking in as image data.
It is input to the change judgment / extraction means 6 to be extracted. A time entry unit for adding the time of occurrence of the abnormal phenomenon in the frame image upon detection of the occurrence of the abnormality and an image format (Join
a frame image editing means 8 having an image conversion unit for automatically converting the image into a photographic expert group (format).
The office monitor 11 has a communication device 9 for transmitting frame images and transmitting monitoring conditions. The change judging / extracting means 6 and the frame image editing means 8 are connected to the office monitoring unit 1.
1 may be included.

A terminal device 12 is installed in the office monitoring unit 11, and the terminal device 12 is connected to the remote monitoring unit 1 and the communication line 10.
A communication device 13 for performing communication via the communication device, a target area setting unit 15 for setting a monitoring target area, and monitoring conditions set by the target area setting unit 15 are transmitted to the image change detecting device 4 in the remote monitoring unit 1. It has a transmitting side remote control means 14 for creating a control signal. The monitoring (attention) area setting means 15 may be provided in the remote monitoring section 1. further,
It has a result display device 16 for displaying a monitoring state, a result storage device 18 for storing monitoring results, and an alarm device 17 for issuing an alarm when an abnormality in the monitoring area is detected. Thereby, the monitoring person converts the attention area setting control by the remote control operation from the attention area setting means 15 on the terminal device 12 by the office monitoring unit 11 into the control signal by the transmission side remote control means 14, and It is transmitted via the device 13, the communication line 10, and the communication device 9 installed in the remote place monitoring unit 1. As a result, it is possible to set a monitoring area that requires detection of an image change or to change the setting. In addition, the monitoring image of the monitoring target area by the imaging device 2 installed in the remote place is not always transmitted to the office monitoring unit 11 in real time, but is obtained by the image change detection method set in the monitoring target area as described above. It is limited to image data at the time of occurrence of an abnormal phenomenon that satisfies the judgment criteria.

However, the still image data for a predetermined period before and after the occurrence of the abnormality is held by using a FIFO register or the like. This is because, when an abnormality occurs, the image data before and after the time including the abnormality is used for investigating the cause.

At this time, the image data transmitted from the remote monitoring unit 1 to the office monitoring unit 11 is a time-series frame image at the moment of occurrence of the abnormal phenomenon in the target area in the monitoring target area. The time at which the abnormal phenomenon occurred is added to the frame image in the time entry section in the image editing means 8 and the file format (Joint photog) which can be used on the terminal in the image conversion section
Still image data converted to a raphic Experts Group format. The transmitted still image data is transmitted to the office monitoring unit 11 as time-series still image data when the image change detection device 4 transmitted from the remote location monitoring unit 1 automatically detects the abnormal phenomenon. As a result of the accumulation, the information is accumulated in the accumulation device 18, and an alarm of the occurrence of an abnormality is generated by the alarm device 17. In other words, the result storage device stores information before and after the occurrence of the abnormality.
Predetermined time-series image data is accumulated.

Next, an outline of the operation of the abnormality monitoring device shown in FIG. 1 will be described with reference to the flowchart of FIG. Photographing device 2
Is started to capture image data. In this case, an attention (monitoring) area is set by an arbitrary line segment from the office monitoring unit 11 as described above. The monitoring target area is set by the monitoring target area setting unit 15 of the office monitoring unit 11, and the transmitting-side remote control unit 14 and the communication device 1
3. A monitoring area of interest is set in the change determination / extraction means 6 via the communication line 10 and the communication device 9 (step 22).
Then, monitoring by the photographing device 2 is started (step 2).
4), images are continuously captured (step 26). Then, it is determined from the state change of the line segment whether or not an abnormality has occurred (step 28). Although the abnormality determination will be described later, the determination is made by paying attention to the line segment that determines the monitoring target area. If it is not determined to be abnormal, the data before Δt with respect to the current time T is deleted (step 30). For example, data from the current time to the time Δt is held using a FIFO register or the like.

Assuming that the time at which an abnormality is determined is T,
The image data continues to be taken only for Δt after the abnormality. That is, as image data, L1 = T−Δt, and L2 = T + Δt.
(Step 32), and transmitted to the result storage device 18 of the office monitoring unit (step 36), to confirm the abnormal state (step 38) or to investigate the detailed cause of the abnormality by taking images before and after the abnormal state. Used for etc. At the same time, if it is determined that there is an abnormality, an alarm is issued by the alarm device 17 (step 34) to notify an abnormal state. Then, it is determined whether or not the monitoring is completed (step 4).
0), end.

Next, the first function will be described. In order to set an arbitrary monitoring area of interest in the monitoring video,
The monitoring target video captured by the capturing device is input to the video input device 5.
Is converted into a digital moving image that can be used on the terminal, captured as a monitoring target video on the image change detection device, displayed using the result display device 16 of the office monitoring unit 11, and displayed on the displayed monitoring video. Then, a monitoring target area is set using a mouse device. The monitoring target area is an area surrounded by a plurality of straight lines or curves with respect to the monitoring target. At the time of this setting, the monitoring target area setting means 15 in the terminal installed in the office monitoring unit via the communication line sets the setting. Monitoring is performed in a state where the above-mentioned monitoring target area is displayed in an overlay manner on the monitoring video. If an object temporarily passes over the set monitoring area of interest, an abnormality is found by detecting the set area of interest based on the determined learning difference. Background subtraction is to take the difference between a background image showing only the background and a frame image containing an intruder (moving object) by using the fact that the background hardly changes in a video taken by a fixed camera. This method is used to separate and extract only intruders (moving objects). However, the abnormality detection here focuses on the line segment in which the monitoring target area is set, and judges an abnormality based on a change in state as the line segment. That is, although the monitoring area is captured as an image, the abnormality determination focuses on a line segment surrounding the monitoring area. Therefore, the data amount can be reduced as compared with the case where the background difference of the entire image is obtained.

Next, the second function will be described. Abnormality in the monitored area of interest in the monitored area (abnormal line segment)
When the image change detecting device detects the occurrence of the image, the detected frame image is displayed on the result display device of the office monitoring unit by the communication device, and at the same time, the frame is input by the time entry unit of the frame image editing means 8. An abnormal phenomenon occurrence time is added to the frame image, and the frame image is converted into an image format (Joint photographic E) that can be used on a terminal by an image conversion unit in the frame image editing means.
xperts Group format), and the converted frame
The system image is stored in the result storage device in the office monitoring unit by the communication device.

Next, the third function will be described. Still image data for a predetermined time before and after the moment of the image change is accumulated so that it can be used as a moving image. That is, moving image data on the terminal (moving image data on the terminal)
The data is temporarily stored in the result storage device 18 as a maximum of 30 frames of still images per second). At this time, the temporary storage time is set to an arbitrarily set time (L), and the moving image data after a predetermined time is automatically deleted by the result output means in the remote place monitoring unit. Next, when the image change detection device automatically detects the occurrence of the image abnormality phenomenon, a certain time (Δt) before the image abnormality phenomenon detection from the moving image temporarily stored for the moment (T) at that moment. The moving image (L1) is extracted from the temporarily stored moving image data of the arbitrary time (L) described above, and a predetermined time (Δ
This is performed by synthesizing the moving image (L2) for t) and storing it in the result storage device in the office monitoring unit via the communication device. In addition, the time for temporary storage is set arbitrarily (L),
It is longer than the video (L1) for a certain time before the detection of the image abnormality phenomenon.

An example in which an exhibit is monitored using the remote location abnormality monitoring device of the present invention will be described with reference to FIGS.

The photographing device is installed above the exhibit S1, and photographs the object to be monitored. Next, A1, A2, A1, A2, and A1 indicate target areas to be automatically monitored for abnormal phenomena by a remote control operation performed by the monitoring area setting unit provided in the office monitoring unit 11.
A no-entry area is set by being surrounded by a square with four straight lines A3 and A4. Then, monitor for abnormalities on this line.

When the hand of the onlooker P2 crosses the boundary line A4 as shown in FIG. 3 in the monitoring target region in the monitoring target region, a change in the state of the boundary line A4 occurs. Automatically detect abnormal changes in the image. P1
Is also a spectator, but is not detected as an abnormal state until it crosses the boundary lines A1 to A4. K1 is an alarm device.

The abnormal line segment will be described with reference to FIG. The abnormality detection by the line segments A1 to A4 is performed by the same method.
Shows the case of the line segment A4 in correspondence with FIG. The line itself is the target of the abnormality judgment, and FIG. 4 (A)
Shows a line segment A4. In FIG. 7B, at time t1, this line segment information is stored in pixel units such as d1t1 to dnt1. Then, the pixel information d1t2 to dnt2 at time t2 is stored as shown in FIG. At this time, for example, if the pixel information is “dit1 ≠ dit2”, it is determined that the line segment A4 has an abnormality. In this case, the pixel unit is described. However, if the pixel size is too small, the determination may be made in a plurality of pixel units. In the example of FIG. 3, pixel information of A4 at t1 and t2 is not equal among a plurality of pixels. A monitoring cycle of about 30 frames / second is sufficient.

In this case, even if a spectator enters the monitoring target area, no change in the state of the boundary line occurs unless the user enters the no-go area (target area), so that no abnormal phenomenon is detected. When the change / judgment / extraction means of the image change detection device automatically detects the abnormal phenomenon, the time-series still image data at that moment is transmitted to the office monitoring unit via the communication line. In an environment where LAN is laid in the exhibition hall due to facilities,
LAN can be used for data transfer between the remote monitoring unit and the office monitoring unit. The alarm display device that senses the reception of the time-series still image data can also emit an alarm sound K1 together with the blinking of the emergency light, and the observer who senses this, as an abnormal phenomenon on the result display device, By confirming the detected event, it is possible to take measures such as preventing dangerous behavior. By installing the alarm display device, the observer can detect the occurrence of abnormal phenomena in the area of interest in the monitoring target area from the alarm sound or the blinking of the emergency light, etc., so the image of the abnormal phenomenon is always displayed This eliminates the need to monitor a TV monitor. Therefore, the monitoring operation can be performed in parallel with other operations, and the efficiency of the monitoring operation can be significantly improved.

The time-series still image data at the moment of occurrence of the abnormal phenomenon is stored in the result storage device, and can be used as a proof image when a criminal act or the like occurs.

[0026]

According to the present invention, it is possible to monitor only a necessary area in a video taken by a surveillance camera when monitoring a remote place. In addition, since time-series image data including image data at the moment of occurrence of the abnormality that is regarded as important is stored, it can be applied to investigation of the cause of the occurrence of abnormality using the stored image.

[Brief description of the drawings]

FIG. 1 is a diagram illustrating a configuration example of a system that implements the present invention.

FIG. 2 is a flowchart illustrating the operation of a system that implements the present invention.

FIG. 3 is an example of monitoring an exhibit using the present invention.

FIG. 4 shows an example of an abnormality occurrence determination by line segment monitoring according to the present invention.

[Explanation of symbols]

DESCRIPTION OF SYMBOLS 1 ... Remote monitoring part, 2 ... Photographing device, 3 ... Terminal device, 4 ...
Image change detection device, 5: video input device, 6: change judgment
Extraction device 7, Result output means 8, Frame image editing means 9, Communication device 10, Communication line 11, Office monitoring unit,
Reference numeral 12: terminal device, 13: communication device, 14: transmitting side remote control device, 15: attention area setting means, 16: result display device, 17: alarm device, 18: result storage device, S1: exhibition, P1: observation Person
1, P2: onlooker, 2A1 to A4: attention area boundary line, K1: alarm generation device

Continued on the front page (51) Int.Cl. ⁷ Identification symbol FI Theme coat II (Reference) H04N 5/907 H04N 5/907 B 5C087 5/915 5/91 K (72) Inventor Yasuhiro Sekine Yukicho, Hitachi-shi, Ibaraki 3-2-1, Hitachi Engineering Co., Ltd. (72) Inventor Yoshitaka Isaka 3-2-1, Yachimachi, Hitachi-shi, Ibaraki F-term within Hitachi Engineering Co., Ltd. F-term (reference) 5C022 AA01 AB65 AC00 AC18 AC31 AC69 5C052 AA17 DD10 GA02 GB01 GB10 GE05 5C053 FA09 FA11 FA27 JA16 KA04 LA01 LA14 5C054 AA01 AA05 CA04 CC03 CH03 DA01 EA01 EA05 EA07 FA00 FC11 GA04 GB01 GD03 HA19 5C084 AA03 AA09 BB04 BB31 GG04 DD03 DD49 EE05 EE06 EE07 FF22 GG02 GG08 GG19 GG23 GG36

Claims (3)

[Claims] In a method of monitoring an object to be monitored by an image captured by a photographing device, a monitoring area of interest surrounding a monitored object image in a photographed image of the monitored object is set by a plurality of line segments, and the line segment is monitored. Monitoring the state of the line repeatedly at a predetermined cycle, and detecting an abnormality based on a change in the state of the line segment. 2. The abnormality monitoring system according to claim 1, wherein the monitoring object photographing screen data is stored for a predetermined period leading to the line segment abnormality and for a predetermined period after the abnormality. Method. 3. An apparatus for monitoring an object to be monitored by an image capturing apparatus, wherein an image input device for inputting captured image data of the object to be monitored, and the object to be monitored is displayed on the input image by a plurality of line segments. An abnormality monitoring device comprising: a surrounding area of interest to be monitored setting means; and a state change determination means for monitoring a state change of the line segment at a predetermined cycle.