JP3886246B2 - Image surveillance system - Google Patents

Description

[0001]
BACKGROUND OF THE INVENTION
The present invention relates to an image monitoring system that detects an abnormality such as an intrusion in a monitoring area by an image sensor and checks an abnormal state by viewing an image when the abnormality occurs in a remote security center.
[0002]
[Prior art]
In a conventional image monitoring system, an image sensor continuously captures a monitoring area, and determines whether or not an abnormality has occurred based on the captured image. When an abnormality is detected, an image at the time of occurrence of the abnormality is transmitted to the security center. At the security center, the controller checks the image of the abnormality that has been sent and confirms whether the abnormality has occurred and the type of abnormality.
[0003]
[Problems to be solved by the invention]
Such an image monitoring system is described, for example, in Japanese Patent Application No. 10-12813 “Image Monitoring System” by the same applicant. However, as such image monitoring systems become common, intruders or their friends intentionally put a cover on the image sensor lens during warning cancellation (for example, during the daytime), and the alarm set mode is set at night. In this case, the capability of the image sensor may be removed in advance, and intrusion may occur on the image sensor. In this case, even if the warning set mode is set at night, the image sensor cannot recognize the change of the image by the intruder due to the lens cover, so it is judged normal and the occurrence of the abnormality cannot be detected and the security center is notified. do not do.
[0004]
Alternatively, the camera position is moved regardless of whether the alarm release mode or the alarm set mode is in effect, and the camera is directed to a position where an intruder is not captured. This is also a category of visual field disturbance.
In normal abnormality detection logic for detecting whether there is an intruder, an image when there is no abnormality is used as a reference image, and the difference between the reference image and the current image is taken to determine the presence of an intruder. Since the occurrence of an abnormality based on an intruder appears locally, it can be easily detected by such a logic that takes the difference. However, when a car headlight is incident on the image sensor, a difference between the reference image and the current image may occur over the majority of the image.
[0005]
Thus, when the area where the difference occurs exceeds a predetermined range, it is not determined that the intrusion abnormality has occurred. However, even in the warning set mode, there is a possibility that the visual field is obstructed by irradiating the image sensor with light or entering from the blind spot of the camera and covering the camera.
The visual field disturbance occurs not only by covering the camera lens with a cover but also by covering the infrared light projecting means of the image sensor.
[0006]
Recently, in order to cope with the situation where the blind spot of such an image monitoring system is struck, a system has been developed in which an image sensor is provided with means for detecting visual field abnormalities that are considered to be visual field disturbances in addition to detecting intrusion abnormalities. ing. In this system, even in the daytime alert release mode, the visual field abnormality detection system is always driven (24-hour monitoring), and when there is visual field obstruction, it is detected immediately and the occurrence of visual field obstruction is reported to the security center.
[0007]
In such a conventional visual field disturbance detection system, the presence / absence of the visual field abnormality is determined based on the same abnormality determination criterion without distinguishing between the warning set mode and the warning release mode. However, during the warning cancellation mode, even if the visual field obstruction action is performed on the camera, the user is in the monitoring area, so the abnormality determination is less urgent. Also, because the user is in the surveillance area, there is a probability of visual field abnormalities that do not directly lead to criminal acts such as putting posters to cover the camera's visual field by mistake or covering the lens for cleaning. high.
[0008]
On the other hand, the occurrence of visual anomaly during the warning set mode means that a person who can execute disturbance, that is, an intruder is in the warning area, and therefore requires urgent action. In addition, since the alarm set mode is normally unattended, the cause of erroneous recognition in the monitoring area is much less than in the alarm release mode. For this reason, even a slight visual field abnormality needs to be immediately determined as abnormal and notified to the security center.
[0009]
Therefore, normally, in order to avoid the misreporting, the detection sensitivity of occurrence of abnormal visual field is increased by using a low determination level during the warning set mode. For this reason, there has been a problem that detection of false alarms, that is, visual anomalies that are not directly related to criminal acts, increases in the warning cancellation mode.
The present invention has been made to solve such a problem in the image monitoring system, and changes the determination level of occurrence of abnormality that is considered to be visual field disturbance during the warning set mode and the warning release mode. It is an object of the present invention to provide an image monitoring system capable of performing the above.
[0010]
[Means for Solving the Problems]
The present invention has been made to achieve the above object. Therefore, the present invention provides an image sensor that images a monitoring area and determines whether or not an abnormality has occurred from the obtained image, and a monitoring area Alert mode And an image monitoring system comprising a controller for setting a warning set mode or a warning release mode. Storage means for storing a determination level for determining whether or not an abnormality has occurred, and a determination level, Alert set mode Set to the first level if Alert release mode Is set to a second level having a lower detection sensitivity than the first level. Judgment level change means And image processing means for determining presence / absence of abnormality using the determination level from an image captured regardless of the warning mode; An image monitoring system is provided.
[0011]
According to this image monitoring system, an appropriate determination according to the security mode of the monitoring area can be performed by appropriately changing the abnormality determination level depending on the monitoring area in the warning set mode or the warning cancellation mode. For example, it is possible to reduce false alarms based on unintentional visual disturbances that are likely to occur in the alert release mode, and to prevent misreports of abnormal occurrences in the alert set mode.
[0012]
Furthermore, the abnormality determination level according to the present invention is the number of feature extraction areas in which an abnormality is detected among a plurality of preset feature extraction areas in a captured image of the image sensor, or the duration of time during which the abnormality is detected by the detection means. Based on the warning set mode or the warning release mode setting in the controller, one or both of the abnormality determination levels are changed.
[0013]
This makes it possible to easily and reliably change the abnormality determination level.
Further, the abnormality determination level changing means of the present invention is configured to monitor in the unattended state by lowering the determination level in the warning set mode than in the warning release mode, that is, by increasing the sensitivity of abnormality detection. Do it strictly.
[0014]
As a result, in the present invention, for example, the judgment level of visual field abnormality occurrence in the warning set mode is lower than in the warning release mode, and as a result, false alarms based on unintentional visual disturbance can be greatly reduced during the warning release mode. In addition, it is possible to prevent misreporting during the warning set mode.
[0015]
DETAILED DESCRIPTION OF THE INVENTION
Embodiments of the present invention will be described with reference to the drawings.
FIG. 1 is a diagram showing an overall configuration of a monitoring system to which an image monitoring system of the present invention is applied.
A controller 1 is installed in a building to be monitored, and an image sensor 2, a fire sensor 3, an emergency button 4, and a mode setting device 5 are connected to the controller 1. The controller 1 is connected to a security center device 7 provided in a remote security center via a telephone line 6 as a communication line.
[0016]
Here, the operation of the entire monitoring system of FIG. 1 will be briefly described.
The mode setting unit 5 sets the monitoring system to a warning release mode or a warning set mode. When the sensor detects an abnormality in the warning set mode, the controller 1 transmits an abnormality signal indicating that an abnormality has occurred to the security center device 7 via the telephone line 6.
[0017]
When an abnormality signal is output from the fire sensor 3 or the emergency button 4, the controller 1 transmits data indicating the type of abnormality and the location where the abnormality has occurred to the security center device 7 through the telephone line 6. Note that these abnormal signal detection methods, transmission methods, and the like are well known in the art, and will not be described further.
A plurality of image sensors 2 are installed in a building, shoot a monitoring area including a monitoring target such as a window and a door, and determine whether there is an abnormality such as an intrusion from the obtained image. Various methods for determining the presence / absence of this abnormality have been proposed, and any method can be employed.
[0018]
In the present invention, the image sensor 2 further includes means for detecting a visual field abnormality of the sensor.
Hereinafter, an example of a technique for detecting an abnormality from an image and an example of a visual field abnormality detection technique will be described with reference to FIG.
2A is a reference image, FIG. 2B is a current image, FIG. 2C is a difference between the reference image and the current image, and FIG. 2D is a diagram illustrating a feature extraction region for visual field abnormality detection.
[0019]
The current image of (B) is the latest image obtained by the image sensor 2, and the difference (C) between the two images is extracted as compared with the reference image of (A). When the difference area between the two images is within a predetermined range, it is determined that an intrusion abnormality has occurred. Here, the predetermined range refers to a range that is wider than a difference region that appears in the case of a small animal such as a mouse and narrower than a difference region that covers a wide range such as irradiation of a headlight of an automobile. That is, in order to distinguish humans from small animals and lighting fluctuations, this is a range that has been previously determined by experience or experiment. In addition, as the reference image of (A), (b) an image obtained when the image sensor 2 is installed, (b) an image obtained when the mode is shifted to the warning set mode, (c) every predetermined time interval (e.g., Images taken every 10 frames), (d) one or more images before the current image, etc. are used.
[0020]
Next, a case where an intrusion abnormality is not recognized and a visual field abnormality is recognized will be described. The determination of the intrusion abnormality is equivalent to, for example, a predetermined pixel (for example, half of all images) in order to prevent the case where the headlight of the automobile is irradiated within the field of view of the image sensor 2 from being mistaken as an intruder. If there is a change in the number of pixels), it is determined not to be abnormal. For this reason, for example, when the intruder does not enter the field of view of the image sensor 2 and is covered by the field of view of the image sensor 2, the field abnormality may be recognized without the intrusion abnormality being recognized. For example, when the intruder approaches the image sensor 2 without entering the field of view of the image sensor 2, for example, the intruder lurks in advance in the blind spot position of the image sensor 2 during the alert release mode, and the alert set mode is set. The case where it covers later is assumed.
[0021]
A method for detecting visual field abnormality will be described.
In one embodiment of the present invention, a plurality of (seven in the case of FIG. 7D) feature extraction regions 8 for visual field disturbance detection are appropriately set in advance on a screen obtained by an image sensor. It is desirable that these areas 8 are not concentrated in one place on the screen as much as possible but are dispersed. In each feature extraction region 8, the difference between the reference image and the current image is taken, and when the difference is equal to or greater than a predetermined value, it is recognized that an abnormality has occurred. As the reference image for visual field disturbance detection, the above-mentioned images (A) and (B) are used. Anomalies based on visual field disturbance usually occur over a wide range, and unlike the occurrence of anomalies due to intrusion, they do not occur locally. Therefore, in this embodiment, it is first determined whether or not an abnormality has been detected in a preset number of feature extraction regions 8. Furthermore, since an abnormality due to visual field disturbance has a relatively long duration, a duration is set in advance, and it is determined whether or not the occurrence of the abnormality continues over this time. That is, the number of monitoring areas 8 recognized as having an abnormality and its duration are used as criteria for visual field abnormality.
[0022]
Further, in the alert release mode, since there is a user in the monitoring area, the visual field abnormality determination level is set to a high level to prevent false alarms. Specifically, when an abnormality occurs in six of the seven feature extraction regions 8 and the duration of the abnormality is 3 minutes or more, it is determined that there is a visual field disturbance. On the other hand, in the warning set mode, since there is no user in the monitoring area, this determination level is kept low, and the detection sensitivity of the occurrence of visual field abnormality is increased to avoid the false alarm. For example, if an abnormality occurs in the three feature extraction regions 8 for one minute or longer, it is determined that an abnormality has occurred due to visual field disturbance.
[0023]
As described above, in the present embodiment, the number of feature extraction regions is determined by the number of feature extraction regions where an abnormality has occurred and whether the state exceeds a certain duration or not as the determination level of occurrence of visual field disturbance. The determination level can be easily changed by changing the duration time by software together with the setting of the warning mode in the controller 1. The abnormality determination level may be changed only for the number of occurrences of abnormality in the feature extraction region 8 or only for the duration of abnormality.
[0024]
Next, the operation of the entire image monitoring system when the image sensor 2 detects a normal intrusion abnormality will be described using the time chart of FIG.
When the mode is changed by the setting of the mode setting device 5, the controller 1 transmits a mode change signal to all the image sensors 2. Further, the controller 1 sequentially sends a status call signal to each of the image sensors 2A, 2B... (Polling).
[0025]
Each of the image sensors 2A, 2B,... Sends a status signal, that is, a normal signal or an abnormal signal to the controller 1 in response to the status call signal.
When the image sensor 2A transmits an abnormal signal in response to the status call signal, the controller 1 dials and connects to the security center device 7 via the telephone line 6. In addition, an abnormal image transmission request signal is transmitted to the image sensor 2A that has transmitted the abnormal signal.
[0026]
The image sensor 2A transmits an abnormal image to the controller 1 in response to the abnormal image transmission request signal. When there is a response from the security center device 7, the controller 1 transmits the abnormality signal sent from the image sensor 2 </ b> A and the image at the time of occurrence of the abnormality to the security center device 7 via the telephone line 6. When this transmission is completed, an abnormal image deletion signal is output to the image sensor 2A, and the image sensor 2A deletes the stored abnormal image and continues the monitoring operation.
[0027]
The security center device 7 displays an abnormal image on the display unit with the reception of the abnormal signal as a trigger, and the controller looks at this image to confirm the abnormal state. When the abnormal state can be confirmed, a reception end signal for the controller 1 is output.
Upon receiving the reception end signal, the controller 1 opens the telephone line 6 connected to the security center device 7 and outputs a status call signal to the next image sensor 2B. Thereafter, the same operation as described above is repeated.
[0028]
Here, the confirmation method of the state in the security center apparatus 7 is demonstrated.
The security center device 7 creates an image when an abnormality occurs from the video signal sent from the controller 1. The controller looks at the image at the time of the occurrence of the abnormality and confirms whether or not the abnormality has actually occurred or what the abnormality has occurred.
Note that the image sent from the image sensor 2 to the security center device 7 can include images of several frames before and after that in addition to the image at the time of occurrence of abnormality. In this case, the controller can improve the accuracy of confirming the abnormal state by comparing the image at the time of occurrence of the abnormality with the images before and after the image.
[0029]
Next, specific configurations of the controller 1, the image sensor 2, and the security center device 7 used in the image monitoring system will be described.
FIG. 4 shows the configuration of the image sensor 2, and FIG. 5 shows the contents of the storage means in the image sensor 2.
The image sensor 2 is provided with a control unit 201 including a CPU and a power source 202. Each part in the sensor is controlled by the control unit 201 and receives power supply from the power source 202. The power source 202 includes a conversion device that converts an AC voltage supplied from the outside into a DC voltage. Note that power may be supplied from the controller 1 instead of providing the power source 202 in the image sensor 2.
[0030]
A storage unit 220 is connected to the control unit 201. The contents of the storage means 220 will be described with reference to FIG.
In the storage means 220, a program area 221 storing a program for causing the image sensor 2 to execute a predetermined operation, a parameter area 222 storing a parameter, a work area 223, and a monitoring area state, that is, an abnormality has occurred. A state storage area 224 for storing the distinction between the normal state and the normal state, and a mode storage area 225 for storing whether the monitoring area is in the alert set mode or the alert release mode.
[0031]
The storage means 220 further includes a reference image storage area 226, a feature extraction area coordinate storage area 227 for visual field abnormality detection, a current image storage area 228, and an abnormal image storage as areas for storing images taken by the imaging means 203. A region 229 is provided. The current image storage area 228 stores the latest image captured by the imaging unit 203. The reference image storage area 226 stores the reference image described with reference to FIG.
[0032]
Note that the reference image may be different from the one for detecting an intruder and the one for detecting visual field disturbance, or the same image may be used. In the abnormal image storage area 229, when an abnormality is detected in the current image, the current image at that time is stored as an image when the abnormality occurs. Alternatively, it may be configured to store a total of 10 frames of the current image, the previous 2 frames, and the subsequent 7 frames.
[0033]
Returning to FIG. 4, the imaging means 203 for photographing the monitoring area is constituted by a CCD camera. Further, an infrared light projecting means 204 is provided, and infrared light is projected to the monitoring area when the mode is shifted to the warning set mode.
The image processing unit 205 compares the current image stored in the storage unit 220 with the reference image to determine whether an abnormality has occurred. The basic idea of this determination has already been described with reference to FIG. The display unit 206 is configured by an LED, and is turned on when an abnormality is detected, and is turned off when an abnormality is not detected. The display unit 206 displays the presence / absence of abnormality detection outside the image sensor 2.
[0034]
The communication unit 207 is an interface that transmits and receives signals to and from the controller 1, and is connected to an image sensor communication control unit 105 (to be described later) in the controller 1 via a signal line 209A. The image output unit 208 is an interface for outputting an image when an abnormality occurs, and is connected to a video input / output control unit 106 (FIG. 6) described later in the controller 1 through a signal line 209B.
[0035]
The address setting portion 210 is configured by a dip switch, and an address for the controller 1 to specify the image sensor 2 is set. The operating unit 211 is a unit for turning on and off the power.
FIG. 6 shows the configuration of the controller 1.
The controller 1 is provided with a control unit 101 and a power supply circuit 102 configured by an MPU or the like. Each part in the controller 1 is controlled by the control unit 101 and receives power supply from the power supply circuit 102. The power supply circuit 102 includes a conversion circuit that converts an external AC voltage into a DC voltage and a battery. A mode setting device 5 is also connected.
[0036]
The controller 1 is further provided with the following parts.
The sensor monitoring circuit 103 is an interface with the fire sensor 3 and the emergency button 4. A modem 104 is provided between the telephone line 6 and the modem 104. The image sensor communication control unit 105 is connected to the image sensor 2 by a signal line 209A. The video input / output control unit 106 is connected to the image sensor 2 by a signal line 209B. The video input / output control unit 106 distributes the video signal sent from the image sensor 2 to the modem 104 and the monitor device 10 connected to the outside.
[0037]
The display unit 107 normally displays the monitoring state of the monitoring area on the screen. When an abnormality is detected, the display unit 107 sounds a buzzer and displays the type of abnormality and the location where the abnormality has occurred on the screen. The display unit 107 is also used when setting the image sensor 2. The setting means 108 is used at the time of initial setting or setting change of the image sensor 2.
The general operation of the controller 1 will be described.
[0038]
At the time of initial setting or setting change of the image sensor 2, the worker connects the monitor device 10 to the video input / output control unit 106 and is displayed on the monitor device 10 in accordance with instructions displayed on the display unit 107 and the monitor device 10. The image sensor 2 is set from the setting unit 108 (for example, height, angle, sensitivity, etc.) while confirming the monitored area. In this way, by setting the image sensors 2 in the controller 1, it is possible to set all the image sensors 2 at one place. The monitor device 10 is removed from the controller 1 after the setting is completed.
[0039]
When the mode setter 5 sets a warning set mode or a warning release mode and a mode shift occurs, a mode shift signal is transmitted to all the image sensors 2 via the image sensor communication control unit 105 and the signal line 209A.
When an abnormal signal is input from the fire sensor 3 and the emergency button 4 via the sensor monitoring circuit 103, the controller 1 sends an abnormality type and an abnormality to the security center device 7 via the modem 104 and the telephone line 6. Data such as the location and time of occurrence is transmitted.
[0040]
The controller 1 transmits a status call signal, a mode transition signal, an abnormal image transmission request signal, an abnormal image deletion signal, and the like to the image sensor 2 via the image sensor communication control unit 105 and the signal line 209A. Receive normal and abnormal signals. The video signal from the image sensor 2 is received via the signal line 209 </ b> B and the video input / output control unit 106.
[0041]
When the image sensor 2 detects an abnormality, the controller 1 transmits the abnormality signal and the video signal received from the image sensor 2 to the security center device 7.
Note that transmission / reception of signals between the controller 1 and the image sensor 2 and between the controller 1 and the security center device 7 has already been described with reference to FIG. 4, so refer to this description.
[0042]
In this example, an analog telephone line is used as a communication line, but a digital telephone line such as ISDN can also be used. In this case, means for converting the video signal from an analog signal to a digital signal is required.
FIG. 7 shows a configuration of a part related to the monitoring system in the security center device 7.
[0043]
The security center device 7 is provided with a control unit 701 configured by an MPU or the like, and each part in the security center device 7 is controlled by the control unit 701.
A communication interface 702 for communicating with the controller 1 through the telephone line 6 is provided. The video signal and the related signal sent from the controller 1 are stored in the image information storage unit 703. The video signal is stored after being converted into digital data. This video signal may include an image at the time of occurrence of an abnormality and images before and after that. The data stored in the image information storage unit 703 includes the address of the image sensor 2 that detected the abnormality, the abnormality occurrence time, the image frame number, and the image data.
[0044]
When the control unit 701 receives the abnormality signal, the control unit 701 displays on the monitor 704 that the abnormality has occurred, the name of the abnormality occurrence property, the abnormality occurrence location, and the like, and displays an image at the time of occurrence of the abnormality. The monitor 704 is connected to the control unit 701 via the image output unit 706.
In the case where the state cannot be sufficiently confirmed only by the image at the time of occurrence of an abnormality, the control unit 701 can display the subsequent frames of the image at the time of occurrence of the abnormality sequentially or jumpingly.
[0045]
Next, the detailed operation of the image sensor 2 will be described with reference to the flowcharts of FIGS.
FIG. 8 shows an operation flowchart of the image sensor 2. The illustrated operation is started when the image sensor 2 is powered on.
In step S11, the image sensor 2 is initialized. As described above, this initial setting is performed in the controller 1, and the image sensor 2 receives data from the controller 1, stores the data in the parameter area 222, and stores the height, angle, and the like of the imaging unit 203. Set based on the data. When the initial setting is completed, the process proceeds to step S12.
[0046]
In step S12, determination of intrusion and visual field abnormality is performed from the captured image. Details of this will be described later with reference to the abnormality determination flowchart of FIG.
When the abnormality determination is finished, it is determined in step S13 whether or not a status call signal has been received from the controller 1. When receiving (yes), the current state stored in the state storage area 224 in step S14, that is, a normal signal or an abnormal signal is transmitted. After the abnormal signal is transmitted, the state storage area 224 is forcibly returned to the normal state. If the status call signal has not been received (no), step S14 is skipped.
[0047]
In step S15, it is determined whether or not an abnormal image request signal has been received from the controller 1. When receiving (yes), the abnormal image stored in the abnormal image storage area 229 in step S16 is transmitted. If the abnormal image request signal has not been received (no), step S16 is skipped.
In step S17, it is determined whether or not a mode transition signal is received from the controller 1. If it is received (yes), it is determined in step S18 whether or not it is a warning set mode. On the other hand, when the mode abnormality signal is not received in step S17 (no), the process returns to step S12 and abnormality determination is performed again.
[0048]
If it is the warning set mode in step S18 (yes), the level for visual field abnormality occurrence determination is set to a low value. Specifically, as described with reference to FIG. 2D, the number N of feature extraction regions 8 for detecting an abnormality is set to N = 3, for example, and the abnormality duration T is set to 1 minute, for example. To do. If it is not the warning set mode, that is, in the warning cancellation mode (no), in step S20, the level for visual field abnormality occurrence determination is set to a high value. Specifically, the number N of feature extraction regions 8 for detecting an abnormality is set to N = 6, for example, and the duration T of the abnormality is set to 3 minutes, for example. Thereafter, the process returns to step S12 and the abnormality determination is performed again. In the present embodiment, both the number N of feature extraction regions 8 and the abnormal duration T are changed and set, but only one of them may be changed and set. The determination parameter for setting the abnormality determination level to “low” or “high” is not limited to the present embodiment, and for example, the ratio of the number of difference pixels in the entire image may be used.
[0049]
FIG. 9 is a flowchart showing details of the abnormality determination step S12. Therefore, the flowchart of FIG. 9 is inserted between (1) and (2) in the flowchart of FIG.
First, in step S21, the current image is stored in the current image storage area 228, and the process proceeds to step S22. In step S22, it is determined whether or not a warning set mode is stored in the mode storage area 225. In the case of the warning set mode (yes), the process proceeds to step S23, and an intrusion abnormality is determined. The determination of the intrusion abnormality is performed by taking the difference between the images stored in the reference image storage area 226 and the current image storage area 228, the details of which are described in detail in the aforementioned Japanese Patent Application No. 10-12813. Therefore, it will not be described in detail here.
[0050]
When the alert release mode is stored in the mode storage area 225 (no in step S22), step S23 is skipped and the intrusion abnormality is not detected.
Next, in step S24, it is determined whether or not the timer flag is on. Normally, intrusion abnormality detection in step S23 is performed at short time intervals, for example, every 0.5 seconds, whereas visual field abnormality detection need not be performed at such short time intervals. Therefore, in this embodiment, the visual field abnormality is detected at intervals of 10 seconds by setting the timer to 10 seconds, for example, and determining whether 10 seconds have elapsed in step S24.
[0051]
Therefore, if the timer flag is not on in step S24 (no), the subsequent steps are skipped, and the process proceeds to step S13 for determining the reception of the status call signal in FIG. If the timer flag is on (yes), the process proceeds to step S25, the timer flag is turned off, and the process proceeds to step S26 for detecting visual field abnormality.
In step S26, it is determined whether or not the difference between the current image and the reference image is a predetermined value or more, and visual field abnormality is detected. As described above with reference to FIG. 2D, this determination is performed by determining whether or not the number of areas in which the abnormality is detected in the feature extraction area 8 is equal to or greater than a predetermined value N. In the embodiment shown in FIG. 8, N is 6 when the monitoring area is in the alert release mode and 3 in the alert set mode.
[0052]
When a visual field abnormality is detected in step S26 (yes), the process proceeds to step S27, the counter A is set to A + 1, and the count is started. This counting is continued until the value of the counter A reaches m (step S28). This count is for determining whether or not the visual field abnormality state continues for a predetermined period T. Continue for a predetermined period First It is considered that there is intentional visual disturbance. As described with reference to FIG. 8, the value of the period T is short in the alert set mode (for example, 1 minute) and long in the alert release mode (for example, 3 minutes). Therefore, the value of m in step S28 is small in the warning set mode and large in the warning release mode.
[0053]
When the value of the counter A reaches m in step S28, the image at that time is stored in the abnormal image storage area 229 in step S29, the abnormal state is stored in the state storage area 224 in step S30, and the counter A of the counter A is stored in step S31. Return the value to 0. If the number N of feature extraction regions 8 in which an abnormality is recognized in step S26 is equal to or less than a predetermined value, steps S28, S29 and S30 are skipped and the process proceeds to step S31.
[0054]
When it is detected whether or not there is a visual field abnormality as described above, the detection information is transmitted to the controller 1 by executing step S13 and subsequent steps in FIG.
FIG. 10 shows an example of a flowchart of timer flag control. A timer is started by turning on the power (step S41), and it is determined in step S42 whether the timer has elapsed, for example, 10 seconds. When the time has elapsed (yes), the timer flag is turned on in step S43, and the timer is reset in step S44. If no in step S42, the timer continues to count.
[0055]
FIG. 11 shows an operation flowchart of the controller. This operation is started by turning on the power.
In step S <b> 51, the controller 1 transmits parameters for initial setting, data related to the security mode, and the like to each image sensor 2. Thereafter, a state calling signal is transmitted to each image sensor 2 in step S52, and a state signal indicating the current state is received from each image sensor 2 in step S53. In step S54, it is determined from the received status signal whether an abnormal signal has been received. When an abnormal signal is received (yes), a communication line is connected to the security center device 7 (step S55), and then an abnormal image is transmitted to the security center device 7 between steps S56 to S60. Since the processing during this time has already been described with reference to FIG. 3, it will not be described repeatedly here.
[0056]
The abnormality display in step S61 indicates that the abnormality is displayed on the display unit 107 of the controller 1.
When an abnormal image is transmitted to the security center device 7 as described above, it is then determined whether or not there has been a mode transition operation in step S62. If there is a transition operation, in step S63, all image sensors 2 are detected. A mode transition signal is transmitted to and the process returns to step S52. If there is no mode transition operation (no), step S63 is skipped and the process returns to step S52 to start polling of the image sensor again.
[0057]
In the present embodiment, the visual field abnormality has been described. However, the present invention is not limited to this, and the determination level varies depending on the warning set mode and the warning release mode, for example, monitoring of a fire abnormality, monitoring of the operation status of equipment, etc. It goes without saying that is also applicable.
[Brief description of the drawings]
FIG. 1 is a diagram showing an overall configuration of a monitoring system to which an image monitoring system of the present invention is applied.
FIG. 2 is a diagram for explaining an example of a technique for detecting an abnormality from an image of an image sensor.
FIG. 3 is a time chart for explaining the operation of the image monitoring system in FIG. 1;
4 is a diagram showing a configuration of the image sensor in FIG. 1. FIG.
5 is a view showing the contents of storage means in the image sensor of FIG. 4;
6 is a diagram showing a configuration of the controller in FIG. 1;
7 is a diagram showing a configuration of the security center device in FIG. 1;
8 is an operation flowchart of the image sensor of FIG.
9 is a flowchart of abnormality determination for the image sensor in FIG. 4;
10 is a flowchart of timer flag control in the flowchart of FIG. 9;
FIG. 11 is a flowchart showing the operation of the controller of FIG. 1;
[Explanation of symbols]
1 ... Controller
101 ... Control unit
102: Power supply circuit
103. Sensor monitoring circuit
104 ... modem
105: Image sensor communication control unit
106: Video input / output control unit
107 ... display section
108: Setting means
2 ... Image sensor
201 ... Control means
202 ... Power supply
203 ... Imaging means
204: Infrared light projection means
205: Image processing means
206: Display means
207 ... Communication means
208: Image output means
209A, B ... signal lines
210 ... Address setting section
211 ... Operating means
220: Storage means
221 ... Program area
222 ... Parameter area
223 ... Work area
224 ... State storage area
225 Mode storage area
226... Standard image storage area
227: Feature extraction area coordinate storage area
228 ... Current image storage area
229. Abnormal image storage area
3 ... Fire sensor
4 ... Emergency button
5 ... Mode setting device
6 ... Telephone line
7 ... Security center equipment
701: Control unit
702 ... Communication interface
703 ... Image information storage unit
704 ... Monitor
705 ... operation unit
706: Image output unit
8 ... Feature extraction area
10 ... Monitor device

Claims (3)

In an image monitoring system, comprising: an image sensor that images a monitoring area and determines whether or not an abnormality has occurred from the obtained image; and a controller that sets a warning mode of the monitoring area to a warning set mode or a warning release mode. ,
Wherein the image sensor,
Storage means for storing a determination level for determining whether or not the abnormality has occurred;
A determination level change that sets the determination level to the first level when the alert set mode is selected, and sets the second level that is lower in detection sensitivity than the first level when the alert release mode is set. Means ,
Image processing means for determining the presence or absence of the abnormality using the determination level from the captured image regardless of the alert mode ;
An image surveillance system comprising: When detecting the visual field disturbance of the image sensor, before SL-size constant level, the number of feature extraction region detected an abnormality of the plurality of preset feature extraction region in the captured image of the image sensor, the abnormality The image monitoring system according to claim 1, wherein the image monitoring system is determined based on the detected duration of the abnormality in the feature extraction region. The determination level changing means changes the determination level by changing either one or both of the number of feature extraction areas in which the abnormality is detected and the duration of the abnormality. Item 3. The image monitoring system according to Item 2.

Images