JP2012048690A - Monitoring apparatus - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide a monitoring apparatus which notifies a person being present near a safe or a storage cabinet that is an object to be monitored, in the state where it is dark in a monitoring area, as a suspicious person who avoids the public eye.SOLUTION: A monitoring apparatus 10 includes control means 47 in which an image captured by an imaging section 20 is processed, a person who approaches a safe or the like in the state where it is dark in a monitoring area or who stays around the safe under the darkness is determined as a suspicious person who plans the robbery of money and other articles, on the basis of a result of tracking by tracking means 44 and a determination result of lightness determination means 45, an output section 50 is controlled and an abnormality signal is notified to a manager or to the outside.

Description

The present invention relates to a monitoring device that monitors a predetermined monitoring area in a building such as a supermarket or a building with various sensors or cameras, and in particular, detects a person who performs suspicious behavior while a person is present in the monitoring area. It relates to a monitoring device.

Conventionally, in a monitoring device that detects a person who enters a monitoring area, it is possible to switch between a security set state and a security release state, and to detect a person who has entered a monitoring area that should have been unattended in the security set state. It is common.

When a conventional monitoring device detects a person in the monitoring area in the security set state, it notifies the security company using a telephone line, and in the security release state, even if the monitoring apparatus detects a person in the monitoring area, Do not report to the company. Further, switching between the security set state and the security release state is usually performed by a manual operation when the user goes out of the monitoring area. That is, when the monitoring area that has been in the security release state is transferred to the security set state, the user confirms that the monitoring area will be unattended, and the user transitions to the security set state by performing user authentication and button operation. . The transition from the security set state to the security release state shifts to the security release state when the user is authenticated within a predetermined time after the user enters the monitoring area. If the user authentication cannot be performed within a predetermined time after entering the monitoring area, it is determined as an intruder, and an abnormality is reported to the security company.

In addition, there is a monitoring device that automatically shifts to the security set state when the monitoring area becomes unmanned in order to prevent the user from forgetting to switch to the security set state. For example, there is a monitoring device that automatically shifts to a guard set state if a dark state continues for a predetermined time after the inside of the monitoring region becomes dark (Patent Document 1).

JP 2007-272595 A

In a case of theft at a store or the like, there is a case where a thief sneaks into a room where a safe is installed and steals money when the surveillance area is in a security release state. Such a case occurs when the user is located in the monitoring area, but is located in another room, not a room in which an important object such as a safe is installed. Since the conventional monitoring device is in a security release state including the room where the safe is installed, even if the human body detection sensor detects the thief, it was not determined to be abnormal. It is also possible to shift only the room in which the safe or the like is installed to the security set state. However, if the security set state is always set, the usability of the room is significantly reduced.

Moreover, even if the unmanned state of the monitoring area is detected and automatically shifted to the guard set state as in Patent Document 1, a predetermined time for determining that the person is unmanned becomes a problem. On the other hand, in the method of Patent Document 1, if a person remains with the light turned off for some reason, such as when the user is extinguished in the monitoring area and is taking a nap, the user is determined to be an intruder. Then, there was a problem of being notified of abnormalities.

By the way, a user of a room usually performs work while making the room sufficiently bright. On the other hand, a thief or the like often works in the vicinity of a safe or an important document storage room while keeping the room dark so that the user does not notice. If you are a regular user of the room, you should turn on the lights.

Therefore, the present invention has an object to provide a monitoring device that detects a person in a dark monitoring area and determines that the person is a suspicious person when it is detected in the vicinity of an important object such as a safe. .

The present invention is a monitoring device that monitors a monitoring object arranged in a monitoring area, and detects a illuminance in the monitoring area and outputs a brightness that is equal to or less than a predetermined illuminance, and a close area of the monitoring target A person detection unit that outputs when the presence of a person is detected, and an abnormality determination unit that determines that the output from the person detection unit is abnormal when the brightness determination unit outputs a predetermined illuminance or less. There is provided a monitoring device characterized by having a control unit that outputs an abnormal signal when the abnormality determining unit determines that an abnormality has occurred.

In such a monitoring apparatus, it is preferable that the abnormality determination unit determines that the person is abnormal if the person is located in the close region of the monitoring target continuously for a predetermined time or more based on the output from the person detection unit.

The monitoring apparatus further includes a storage unit that stores in advance the position of the monitoring object in the monitoring area, and a tracking unit that tracks a person existing in the monitoring area, and the abnormality determination unit includes a brightness determination unit. It is preferable to determine that the person is close to the monitoring target when the output is less than or equal to the predetermined illuminance.

The present invention is a monitoring device that monitors a monitoring object arranged in a monitoring area, and includes a brightness determination unit that detects an illuminance in the monitoring area and outputs a value that is equal to or lower than a predetermined illuminance, and a monitoring area. It has a tracking unit that tracks a person, and when the brightness determination unit outputs that the illumination intensity is less than or equal to a predetermined illuminance, it is determined that the tracking unit is abnormal if the person is wandering in the monitoring area for a predetermined time And a control unit that outputs an abnormality signal when the abnormality determination unit determines that there is an abnormality.

In such a monitoring device, the brightness determination unit determines the illuminance from the average luminance of the image captured in the monitoring area, and the person detection unit extracts the person from the image and extracts the person from the extraction position to the close area to be monitored. The presence is detected, and the tracking unit preferably tracks the person from images obtained by sequentially capturing the monitoring areas.

As a suspicious person who tries to take away a piece of money by avoiding a person who approaches an important object with the light turned off by installing the monitoring device according to the present invention in a monitoring area where the light is normally turned on. Can detect and report.

As one embodiment, it is a schematic diagram showing a state in which a monitoring device according to the device of the present invention is installed in an office of a store. It is a schematic block diagram of the monitoring apparatus concerning 1st embodiment. It is a figure which shows the detail of the information memorize | stored in a memory | storage part. It is a schematic diagram which sets the near field of the safe which is an important thing. 1 is an overall flowchart of a monitoring device according to a first embodiment. It is a flowchart concerning detection and determination of a suspicious person. It is a schematic diagram concerning a suspicious person's determination. It is a schematic block diagram of the monitoring apparatus concerning 2nd embodiment. It is a whole flowchart of the monitoring apparatus concerning 2nd embodiment. It is a schematic block diagram of the monitoring apparatus concerning 3rd embodiment. It is another schematic diagram concerning determination of a suspicious person.

Hereinafter, as an embodiment of a monitoring apparatus according to the present invention, a case where the monitoring apparatus is installed in an office of a store such as a supermarket will be described as an example with reference to the drawings. Assume that a safe is installed in the office as an example of an important object to be monitored.
During the opening hours when security is canceled, the store manager can enter the office to operate the safe to store sales, as well as general employees and part-time workers can enter the office to perform various tasks. It is.

Here, while security is being released, the office room is basically free to enter and exit, and since there are people, including employees in charge of office work, the office room lighting is usually on.
However, there are cases where all people leave the room due to business reasons and the office becomes unattended. When the person goes unattended, the room lights are turned off to save power consumption. It may be set in the security set state each time, but it is not always required for the convenience of business operation of the store, and the office room may be dark and unattended in the security release state.

A person who enters the office in that state is usually considered to turn on the room lighting. However, it is usually difficult to think of entering the room without turning on the light, that is, the office room is dark, and approaching an important safe. Such a person can be said to be a suspicious person, such as trying to take away the money stored in the safe.
Furthermore, it can be said that the person is more suspicious if he stays in the close range where he can reach the safe and reach the safe without staying close to the safe.

Therefore, if a person who approaches an important object while the office room that is the monitoring area is dark or a person who stays in the immediate area of the important object for a certain period of time is detected, it is reported that a burglary or theft is foreseen.

In the monitoring apparatus according to the present invention, the position of the person in the monitoring area is important information.
In the first embodiment described below, a description will be given assuming that a camera with a substantially vertical downward is installed on the ceiling of the office, and a wide-angle lens that can secure an angle of view over which the entire office can be viewed. The office is the monitoring area.
By processing images sequentially obtained from such a camera in time series, it is possible to detect a person approaching or staying in front of the safe by referring to a preset position of the safe. .

FIG. 1 is a schematic diagram showing a state in which a monitoring device according to the present invention is installed in an office. Examples of the installation location include an office room and a backyard having a storage room in which products before being displayed at a store are arranged on a shelf.
In FIG. 1, a safe 1 is installed in the corner of the floor 3 of the office. Reference numeral 2 indicated by a dotted line is an area where the safe 1 can be operated, and is set on the condition that the safe 1 is at a close distance. In the present embodiment, this is referred to as a close region. This is area information held inside the monitoring apparatus, and is not actually drawn on the floor 3.

The camera 4 is installed substantially vertically downward on the ceiling of the office, and an image acquired from the camera 4 is used for tracking a person existing in the office. It is also used to detect the brightness of the office. An image acquired from the camera 4 is referred to as an input image in the present embodiment. The camera 4 constitutes an imaging unit that will be described later with reference to a schematic configuration diagram of the monitoring device.

FIG. 2 is a schematic configuration diagram of the monitoring apparatus 10 according to the present invention as one embodiment. The monitoring device 10 includes an imaging unit 20, a storage unit 30, an image signal processing unit 40, and an output unit 50.

The imaging unit 20 forms an image of a monitoring area on a two-dimensional detector composed of a photoelectric converter having sensitivity to visible light and near infrared light, such as CCD or C-MOS, and the two-dimensional detector. And a near-infrared illumination 22 that emits light in a wavelength band in which the camera 21 has sufficient sensitivity.

As the camera 21, for example, a camera that continuously shoots according to the NTSC standard can be used. Alternatively, the camera 21 may generate a higher resolution image such as so-called high vision. In this embodiment, the camera 21 generates an image obtained by photographing the office as, for example, a grayscale image in which the luminance of each pixel is represented by 256 gradations.
The camera 21 in FIG. 2 corresponds to the camera 4 in the schematic diagram shown in FIG.

The imaging unit 20 is connected to an interface unit (not shown) of the image signal processing unit 40, generates an image every predetermined time, and outputs the generated image to the image signal processing unit 40 as an input image.

The near-infrared illumination 22 of the imaging unit 20 is turned on when a lighting signal is received via an interface unit (not shown) when a brightness determination unit 45 described later determines that the office serving as a monitoring area is dark. . Similarly, when it is determined that the office is bright, it is turned off in response to a turn-off signal.
Alternatively, if the camera 21 is equipped with a function for determining the brightness of an image and has a contact output terminal, it may be switched on and off by a make break. The near infrared illumination 22 makes it possible to detect a person regardless of the brightness of the office.

In addition, the camera used for the imaging unit 20 may be a type that can acquire not only a near-infrared image but also a color image at the same time.
The camera 21 preferably uses a lens with a wide angle of view, such as an ultra-wide-angle lens or a fish-eye lens, in order to keep the entire office, which is the monitoring area, in view.

The storage unit 30 includes a storage device such as a nonvolatile semiconductor memory such as a flash memory, a volatile semiconductor memory, or a magnetic disk (HDD).
The storage unit 30 stores various programs and data used in the monitoring device 10. The imaging conditions of the imaging unit 20, that is, information such as the height and the angle of view of the imaging unit 20 from the floor are also stored.

In addition, the storage unit 30 stores area information 31, a background image 32, and tracking information 33.
The area information 31 is stored as area information on the input image for an area that can reach the safe 1 that is a monitoring target. The operator having the administrator authority of the monitoring apparatus 10 inputs the near field indicated by reference numeral 2 in FIG. 1 by referring to the input image using an input unit and an image display unit (not shown). The input result is converted into coordinate information by the area setting means 41 and stored.

The background image 32 stores an input image acquired under the condition that no person is captured. It is used to extract an image of a person existing in the office by difference processing with the input image. The background image 32 is created by background image creation means 42 described later.

The tracking information 33 stores various types of tracking information at each time point in the input image of a person extracted by the person extracting unit 43 and tracked in time series by the tracking unit 44, as will be described later. It is.
As shown in FIG. 3A, the tracking information 33 includes a person image number 330, a person image number 331, trajectory information 332, a feature amount 333, an approach counter 334, and a stay counter 335.

The number of person images 330 is the total number of persons currently extracted by the person extraction unit 43 and tracked by the tracking unit 44.
The person image number 331 is a number assigned to each person extracted by the person extracting unit 43 and tracked by the tracking unit 44.
The trajectory information 332 stores coordinate information in the input image from the time when the person with the individual image number 331 appears in the input image to the current time, and each time point. In consideration of the balance with the capacity, the maximum time T is stored.

An example of the format of the trajectory information 332 is shown in FIG. The past time 3320 “0” means the current time, and the coordinates of the barycentric position, which is the position in the input image, are stored in the X coordinate 3321 and the Y coordinate 3322 for the human image obtained from the current input image. To do. The past time point 3320 “1” means the previous time point, and the past time point “T” means the oldest time point.
When new information about the person image is obtained, each coordinate information is sequentially replaced with the one at the past one time point, and the latest coordinate information is stored as the one at the past time point “0”.
The position of the person image is not limited to the position of the center of gravity, but may be coordinates of the vertex of the circumscribed rectangle, coordinates of the center of the circumscribed circle, or coordinates that can be determined as the feet of the person.

The feature amount 333 is a feature amount related to image information about the person of each person image number 331, and is stored by appropriately adopting known feature amounts such as a luminance value histogram, size, shape, aspect ratio, and edge information.
The stay counter 335 holds the number of time points when the person with the person image number 331 is present in the closest region in the input image.
The approach counter 334 holds the number of points in time when the person of each person image number 331 is moving in the direction of the safe 1 in the input image.

Returning to FIG. 2, the image signal processing unit 40 uses the input image input from the imaging unit 20, refers to the information stored in the storage unit 30, tracks a person existing in the office, Processes such as brightness determination and alarm signal transmission.
The image signal processing unit 40 includes an area setting unit 41, a background image creation unit 42, a person extraction unit 43, a tracking unit 44, a brightness determination unit 45, an abnormality determination unit 46, and a control unit 47. These are realized by program modules stored in the storage unit 30.

In the first embodiment, the imaging unit 20 and the brightness determination unit 45 constitute a brightness determination unit in the claims, and the person extraction unit 43 and the area information 31 include the person detection in the claims. The tracking unit 44 and the tracking information 33 constitute a tracking unit in the scope of claims.

The area setting means 41 uses input means and image display means (not shown) to convert the information of the close area 2 for the safe 1 input by a person having administrator authority of the monitoring device 10 into coordinate information in the input image, The area information 31 of the storage unit 30 is stored.

A state of setting the area will be described with reference to FIG. Reference numeral 5 denotes an input image acquired by the imaging unit 20 and shows a state where the office is looked down from the ceiling. Reference numeral 6 denotes a close region designated by a person having administrator authority of the monitoring apparatus 10 and corresponds to reference numeral 2 in FIG. The area setting means 41 converts the area information of reference numeral 6 into coordinate information in the input image and stores it as area information 31 in the storage unit 30.

As the input unit and the image display unit, known ones may be used as appropriate. For example, a mouse that can be externally connected to the monitoring apparatus 10 and a portable small-sized liquid crystal monitor may be used, and an area may be designated by operating the mouse so that the operator includes the close area 2 while displaying an input image.

As described above, the background image creating means 42 uses a so-called background using an input image acquired when the monitoring apparatus 10 is activated or periodically acquired from the imaging unit 20 under the condition that no person is captured. An image is created and stored in the background image 32 of the storage unit 30.
As a method for creating the background image 32, a known method may be adopted as appropriate. For example, the input image may be periodically re-stored in the background image 32 or may be created by averaging a plurality of input images. However, the background image can be kept up-to-date because it can cope with variations in brightness. 32 shall be updated.

The person extraction unit 43 obtains a difference for each pixel between the background image stored in the storage unit 30 and the input image acquired at the present time, and binarizes the difference with a predetermined threshold value. As a result, a change area that satisfies human characteristics such as size, shape, circumscribed rectangle aspect ratio, and the like is specified, and an internal image is extracted as a human image. Further, the center of gravity position (coordinate value) in the input image of the change area that satisfies human characteristics is output to the tracking means 44 as the current tracking position.

Since the input image is an image looked down substantially vertically downward from the vicinity of the ceiling of the office as exemplified by reference numeral 5 in FIG. 4, the person image extracted by the person extracting means 43 is the top and both shoulders. The shape is as seen from above.
The background extraction process generally performed in the field of image processing technology is sufficient for the process performed by the person extraction unit 43, and thus the details are omitted.

The tracking unit 44 associates the tracking information 33 relating to each person image extracted by the person extracting unit 43 up to the previous time point and stored in the storage unit 30 with each person image extracted by the person extracting unit 43 at the present time. In this way, the person image tracking is realized.
For this purpose, the tracking unit 44 extracts feature information such as a brightness histogram and texture information from the person image extracted by the person extracting unit 43 at the present time. And it compares with the feature-value 333 linked | related with the tracking information 33 of each person image number 331 memorize | stored in the memory | storage part 30. FIG.

Furthermore, in consideration of the condition that the position of the center of gravity of the human image at the present time and the trajectory information 332 at the past time stored in the storage unit 30 are close to some extent, the human image number 331 determined to be most similar. Associate with.
The tracking unit 44 updates the trajectory information 332 and the feature value 333 of the associated person image number 331 with the current center of gravity and the feature value.

If the person image extracted at the present time is not associated with any of the tracking information 33 stored in the storage unit 30, the person image newly appears at the current time, that is, the person who entered the office It is determined that
If there is tracking information 33 stored in the storage unit 30 that is not associated with the person image extracted at the present time, the tracking information 33 has moved out of the field of view of the input image. Is determined to have gone out of the office, and is erased from the storage unit 30. Or it is good also as what keeps fixed time and tries said correlation process in the meantime.
The processing in the tracking unit 44 described above is a general tracking process in the field of image processing technology, and thus the details are omitted.

The brightness determination unit 45 processes the input image obtained from the imaging unit 20 and determines the brightness of the office that is the monitoring area. Therefore, the brightness determination means 45 obtains the average brightness of the entire input image, and when it is less than a predetermined brightness threshold, the office room determines that the room lighting is turned off and the room is dark. When it is determined that the office is dark, a signal for turning on the near-infrared illumination 22 is output through an interface unit (not shown), and a signal indicating that is also output to the abnormality determination means 46.
The average luminance only in the vicinity of the safe 1 may be obtained by referring to the area information 31 stored in the storage unit 30 instead of the average luminance of the entire input image.
Furthermore, when the camera of the imaging unit 20 has a brightness detection function, the information may be used as it is.

The abnormality determination unit 46 determines whether or not an abnormal state has occurred in the office based on the region information 31, the tracking information 33, and the determination result of the brightness determination unit 45, and controls the abnormality signal if it has occurred. It outputs to the means 47.
Specifically, when a person who is moving to the safe while the office is dark is detected, or a person who is staying in the nearby area for a predetermined time or longer (for example, 30 seconds or more) while the office is dark is detected The abnormal state is determined and an abnormal signal is output to the control means 47.

The control unit 47 refers to the output result of the abnormality determination unit 46 and controls the output unit 50 to determine whether or not to output an alarm signal. Further, the input image may be stored in the storage unit 30 for the purpose of ensuring trailability.

Next, the operation of the monitoring apparatus 10 according to the first embodiment will be described with reference to FIGS.
FIG. 5 is a main flowchart of the operation of the monitoring device 10.
First, when the monitoring device 10 is activated, initialization is performed (step S10). In this step, one input image is acquired from the imaging unit 20, and the close region 2 is set by the region setting unit 41. An example of the image displayed at this time and the close-up area input by the operator are as shown in FIG.

Next, in step S20, the monitoring apparatus 10 creates a background image 32. For this purpose, the background image creating means 42 acquires a plurality of, for example, 10 input images from the imaging unit 20, obtains the average of the pixel values for each corresponding pixel, and uses the background image 32 of the storage unit 30 as the background image. To remember.

Steps S30 to S80 described below are repeated each time the imaging unit 20 acquires a single image. Each time the imaging unit 20 acquires one image, the time advances by one point.
In step S <b> 30, the image signal processing unit 40 acquires one current input image from the imaging unit 20.

In step S40, the person extraction unit 43 reads the background image 32 in the storage unit 30, obtains a difference for each pixel, and binarizes with a predetermined threshold value, thereby extracting a change area.

Next, the person extracting unit 43 extracts the same type of feature information as the feature quantity 333 that is a part of the tracking information 33 of the storage unit 30 such as the size and the aspect ratio for each extracted difference area. If the extracted feature information matches a predetermined shooting condition or the like, the change area is regarded as being extracted when a person enters the office, and the inside is extracted as a person image. Otherwise, it is determined to be caused by some cause, that is, noise, and is not subjected to subsequent processing.
At the same time, the person extracting means 43 obtains the position of the center of gravity representing the position of the change area determined to be due to the person, and stores the number of extracted person images in the number of person images 330 of the storage unit 30. .
The person extracting unit 43 outputs the obtained information to the tracking unit 44.

Next, in step S <b> 50, the tracking unit 44 reads the tracking information 33 from the storage unit 30 and compares it with the corresponding information regarding the change area input from the person extracting unit 43.
For example, for the feature amount 333, the Mahalanobis distance is obtained for each type of feature amount, and the previous position and distance of the center of gravity position of the change region and the trajectory information 332 of the tracking information 33 are obtained. The tracking unit 44 comprehensively determines them, and determines which of the tracking information 33 stored in the storage unit 30 is most similar to each difference area.

The tracking unit 44 updates the trajectory information 332 of the most similar person image number 331 at the barycentric position of the change area, and also updates the feature quantity 333 with the feature information obtained from the difference area. The update of the trajectory information 332 here is to add the latest barycentric position to the trajectory information 332 stored in the list format, as shown in FIG.

When there is no stored tracking information 33, that is, when the number of human images is 0, or when the current position of the center of gravity of the difference area is more than a predetermined distance from the latest position information of any trajectory information 332 Assuming that the person has newly entered the office, information corresponding to the trajectory information 332 and the feature quantity 333 corresponding to the number “1” of the person image number 331 is stored.

The above processing is repeated for the number of person images extracted by the person extracting means 43. If it is 0, no particular processing is performed, and the process returns to step S30 to acquire the input image at the next time point.

In step S60, the brightness determination unit 45 determines the brightness of the office by obtaining an average luminance value of the entire screen for the input image acquired by the imaging unit 20. When the average luminance of the input image is less than a predetermined threshold, the office is determined to be dark, and when the average luminance of the input image is equal to or higher than the predetermined threshold, the office is determined to be bright.
The determination result of the brightness determination unit 45 is output to the abnormality determination unit 46. At the same time, if it is determined that the office is dark, the near infrared illumination 22 is turned on.

Next, in step S70, the abnormality determination unit 46 refers to the determination result of the brightness determination unit 45 and the tracking result of the tracking unit 44 to determine whether or not an abnormal state has occurred in the office, that is, the office work. When a room approaches the safe 1 while the room is dark, or a person who stays in front (is staying) is detected, a process of determining an abnormal state is performed. Details of the operation of the abnormality determination means 46 in this step will be described with reference to FIG.

First, in step S <b> 700, the abnormality determination unit 46 reads the trajectory information 332, the stay counter 335, and the approach counter 334 for the tracking information 33 of each person image stored in the storage unit 30.
As illustrated in FIG. 3B, the trajectory information 332 is stored in association with a past time point 3320, an X coordinate 3321, and a Y coordinate 3322 for each person image number 331.

In step S705, the abnormality determination unit 46 obtains the direction in which the person image of interest has moved on the input image from the current coordinate information and the previous coordinate information.
In FIG. 3B, the abnormality determination means 46 has the current X coordinate indicated by reference numeral 3323, the current Y coordinate indicated by reference numeral 3324, the X coordinate at the previous time indicated by reference numeral 3325, and the previous time indicated by reference numeral 3326. The movement locus is obtained using the Y coordinate.

This will be described with reference to FIG. In FIG. 7, reference numeral 700 is the center of gravity position in the input image of the person of interest at the previous time, reference numeral 701 is also the current center of gravity position, and reference numeral 702 is the movement locus of the person of interest, which is drawn as a vector.
The abnormality determination unit 46 extends the movement locus 702 from the position 700 at the previous time point in the direction of the current position 701 as indicated by reference numeral 703, and checks whether or not there is a close region 6 (step S710). ).

As shown in FIG. 7, when the close area 6 exists at the extended point, it is determined that the person of interest is approaching the safe, and the abnormality determination unit 46 increments the approach counter 334 of the corresponding person by 1 (step S715). Otherwise, the approach counter 334 is set to 0 (step S720).

Next, the abnormality determination unit 46 refers to the area information 31 in the storage unit 30 and determines whether or not the person of interest is located inside the close area 2 (step S730).
If the person of interest is located inside the close-up area 2, the stay counter 335 is incremented by 1 (step S735). Otherwise, the stay counter 335 is set to 0 (step S740).

Then, the abnormality determination unit 46 checks whether any of the approach counter 334 and the stay counter 335 is equal to or greater than a predetermined threshold (step S750), and when any of them exceeds the threshold, the control unit 47. An abnormal signal is output to (step S755). If none of the threshold values is exceeded, the process moves to step S760, and the next person image is set as a processing target.
Note that if an abnormality signal is output for any person image, the entire abnormality determination process shown in FIG. 6 may be terminated.

Note that the approach counter 334 being equal to or greater than a predetermined threshold means that the person of interest continues to face the safe. Further, the staying counter 335 being equal to or greater than a predetermined threshold means that the person of interest continues to stay in the immediate area of the safe.
Here, as an action of a suspicious person, when entering the office and going straight to the safe, both the approach counter 334 and the stay counter 335 exceed predetermined threshold values set in advance, respectively. Since it may be possible to stay in the immediate area of the safe after this, it is desirable to make an abnormality determination when either the approach counter 334 or the stay counter 335 exceeds a predetermined threshold.

Returning to FIG. 5, when the abnormality determination unit 46 outputs an abnormality signal to the control unit 47 in step S <b> 80, the control unit 47 performs a predetermined output process on the output unit 50. For example, an abnormal signal is transmitted to an external security center using a communication line (not shown), or an e-mail is transmitted to a mobile phone carried by a store manager. Alternatively, the input image can be stored in the storage unit 30 to ensure trail performance.

In the first embodiment according to the present invention described above, the position of the office room and the brightness of the office room are obtained from the imaging unit 20 for the person existing in the office room. The input image was being processed.
However, the present invention is not limited to this. That is, a ranging sensor may be used for detecting the position of a person, and an illuminance sensor may be used for detecting brightness. As the distance measuring sensor, a sensor using microwaves, a sensor using ultrasonic waves or laser light, and the like are well known.

Hereinafter, as a second embodiment according to the present invention, a case where a microwave distance measuring sensor is used to detect the position of a person will be described. In addition, it demonstrates centering on a different part from the method using the camera which is 1st Embodiment described so far, and a common location is simplified and demonstrated suitably or it abbreviate | omits.

FIG. 8 is a schematic explanatory diagram of the monitoring device 100 according to the second embodiment. 2 differs from FIG. 2 in that the position information acquisition unit 200 and the brightness acquisition unit 210 instead of the imaging unit 20, the background information 320 instead of the background image 32, and the background information generation unit 420 instead of the background image generation unit 42. The position information processing unit 400 is provided in place of the image signal processing unit 40.
In addition, elements having substantially the same functions as those in the first embodiment have the same names and are given different symbols.

In the second embodiment, the brightness acquisition unit 210 and the brightness determination unit 450 constitute a brightness determination unit in the appended claims, and the position information acquisition unit 200, the person extraction unit 430, and the region information 310. Constitutes a person detecting section in the scope of claims, and the tracking means 440 and the tracking information 330 constitute a tracking section in the scope of claims.

The position information acquisition unit 200 uses a microwave sensor in the second embodiment, and acquires the distance from the installation position to the object. The installation position is a position where distance information in the horizontal plane direction can be acquired in the office. If only one position information acquisition unit 200 is installed, it is desirable to install it in the vicinity of the safe and toward the center of the office so that the distance to the safe can be directly measured.

If a radar type sensor is used for the position information acquisition unit 200, the horizontal position as the scanning direction can be set to detect not only the distance to the object but also the horizontal position. It is.
When the position information acquisition unit 200 is not a radar type sensor, but a sensor that easily acquires only the distance is used, for example, if three or more sensors are installed with the center direction of the office facing the corner of the office, three points can be obtained. The horizontal position in the office can be detected by the principle of surveying.

The position information acquisition unit 200 obtains a distance to an object existing in the office room every predetermined time and sequentially outputs the distance to the position information processing unit 400. This predetermined time interval corresponds to an image acquisition interval when the camera 21 is used, that is, a so-called frame interval. The interval at which the information is output to the position information processing unit 400 may be determined in consideration of the hardware performance of the specific position information acquisition unit 200 and the like.

The brightness acquisition unit 210 may use a general illuminance sensor and measures the brightness of the entire office. Alternatively, the brightness of only the vicinity of the safe may be measured by appropriately adjusting the optical system. Alternatively, if the office lighting switch ON / OFF information can be acquired, it may be input directly to the position information processing unit 400. This is because it can be determined that the office is bright when the lighting switch is ON and dark when it is OFF. However, depending on the lighting conditions of the office, the office may be sufficiently bright even if the lighting switch is OFF. Therefore, it is desirable to keep the information on the lighting switch as auxiliary information.
The brightness acquisition unit 210 outputs brightness information to the position information processing unit 400 every predetermined time.

The area information 310 of the storage unit 30 is stored in a format that can identify the close area 2 of the safe 1 in the office using the measurement result of the position information acquisition unit 200. Similar to the first embodiment, it is set by the area setting means 41.
For example, the area information 310 is stored as a distance within a predetermined range from the position information acquisition unit 200. Alternatively, the distance information may be converted into a distance image and stored. In that case, processing similar to that of the first embodiment can be performed.

The background information 320 stores a measurement result obtained from the position information acquisition unit 200 under the condition that the office is unattended. For example, when a shelf is placed in the office, it becomes background information including distance information to the shelf. As in the case of image processing, the background information 320 is used to detect a person existing in the office by obtaining a difference from input information from the current position information acquisition unit 200.
If the background information 320 is also converted into a distance image and stored in the same manner as the area information 310, the same processing as in the first embodiment can be performed.

The position information processing unit 400 refers to the input from the position information acquisition unit 200 and the brightness acquisition unit 210 and the information stored in the storage unit 30, and outputs an alarm signal to the outside if a predetermined abnormal state occurs. Has function. This corresponds to the image signal processing unit 40 in FIG.

The background information creation means 410 averages the distance information at a plurality of time points acquired by the position information acquisition unit 200 when the monitoring apparatus 100 is started up or under the condition that the office is unmanned, and the background information 320 is obtained. create. Moreover, it shall update to the newest information suitably. As described above, it may be converted into a distance image.

The person extraction unit 430 obtains the distance to the object based on the background information 320 of the storage unit 30 and the latest distance information input from the position information acquisition unit 200, performs threshold processing under appropriate conditions, and extracts a person. I do.
If the background information 320 is converted into a distance image and stored, and the person extraction means 430 also converts the latest distance information obtained from the position information acquisition unit 200 as a pre-process into a distance image, the first is performed. A person can be extracted by background difference processing as in the embodiment.

Based on the distance information to the person extracted by the person extracting unit 430, the tracking unit 440 tracks a person who satisfies a condition that can be regarded as the same person at each time point by a known method.

The brightness determination unit 450 determines whether the office is bright or dark based on the brightness information of the office acquired by the brightness acquisition unit 210. If the information input from the brightness acquisition unit 210 is, for example, a voltage value, a threshold having a voltage dimension, and if brightness information that has already been A / D converted is input from the brightness acquisition unit 210, A threshold for the value is defined, and the brightness of the office is determined.

The abnormality determination unit 460 determines an abnormality using the region information 310, the tracking information 330, and the input from the brightness determination unit 450. For example, when the brightness determination means 450 determines that the office is dark and the person being tracked by the tracking means 440 continues to move in the direction of the closest area 2 or the office is dark If it is determined that the person has stayed within the close range 2 for a predetermined time or more, an abnormal signal is output to the control means 470 as an abnormal condition has occurred.

FIG. 9 shows an operation flow of the monitoring apparatus 100 according to the second embodiment. This is almost the same as the operation flow according to the first embodiment shown in FIG. 5, except that background information creation (step S200) is substituted for background image creation (step S20), and input image acquisition (step S30) is substituted for sensor. This is information acquisition (step S300). The step having the same name as that in FIG. 5 as the first embodiment is a change point that processes the distance information instead of processing the image information, and therefore the description will be simplified or omitted as appropriate.

These processing contents are almost the same as those of the first embodiment, and distance information is used instead of image information, and processing suitable for the processing may be performed. In particular, if the distance information is converted into a distance image, an abnormality can be determined by the processing according to the first embodiment. Hereinafter, a case where distance information is converted into a distance image will be described as an example.

In step S100, distance information for one time point is acquired from the position information acquisition unit 200, converted into a distance image overlooking the office, and the close-up area 2 is set as in the first embodiment.
In the background information creation process of step S200, the background information creation unit 410 converts the distance information acquired from the position information acquisition unit 200 into a distance image and stores it in the background information 320 of the storage unit 30.

In step S400, the person extraction unit 430 converts the latest distance information acquired from the position information acquisition unit 200 in step S300 into a distance image, and extracts a person by obtaining a difference from the background information 320. At this time, the information necessary for the tracking process is extracted in the same manner as in the first embodiment.
In step S500, the tracking unit 440 performs tracking processing related to the person acquired in the previous step. The flow of processing is the same as in the first embodiment.
In step S600, the brightness determination unit 450 determines the brightness of the office with reference to the output result of the brightness acquisition unit 210. The details of the determination are as described in the description of the brightness determination means 450.

In step S700, the abnormality determination unit 460 determines whether an abnormal state has occurred in the office. That is, when the brightness determination unit 450 determines that the office is dark and the detected person is approaching the safe 1 or stays in the close area 2, the abnormal state is detected. Determine the occurrence.

Since the distance information acquired by the position information acquisition unit 200 is converted into a distance image, it is only necessary to determine approach to the safe or stay in the closest area by the same method as in the first embodiment.
When the distance information is not converted into a distance image, the position information processing unit 400 may define and process a world coordinate system.

Other processes are the same as those in the first embodiment or are not considered necessary for those who have ordinary technical knowledge in the field of monitoring technology, and thus detailed explanations are omitted.

Next, a third embodiment according to the present invention will be described. In the first embodiment and the second embodiment, the tracking process for the human image in the input image has an important meaning. However, in the third embodiment, the configuration is simplified, Only the vicinity is targeted for detection.

FIG. 10 shows a schematic configuration diagram of a monitoring apparatus 1000 according to the third embodiment. In the present embodiment, the monitoring apparatus 1000 is replaced with the person detection unit 250 and the position information processing unit 400 instead of the position information acquisition unit 200, as compared with FIG. 8 which is a schematic configuration diagram according to the second embodiment. Has a signal processing unit 4000. Other components having substantially the same functions as those shown in FIG. 8 are given the same names and different symbols, and the description will be simplified or omitted below.
It should be noted that the brightness acquisition unit 210 and the brightness determination unit 4500 constitute a brightness determination unit in the scope of claims as in the second embodiment.

The person detection unit 250 is a sensor having a function of detecting a person and is not particularly limited. As described above, an image sensor or a microwave sensor may be used. Alternatively, a passive infrared sensor (PIR) or an ultrasonic sensor may be used. However, in any sensor, adjustment is made so that the detection range includes the closest region 6 as necessary and sufficient. As a result, the person detection unit 250 can detect only the person existing in the close area of the safe.

Abnormality determination means 4600 is a case where the office is receiving an output indicating that the office is dark from brightness determination means 4500, and person detection unit 250 outputs that there is a person in the immediate area of the safe. In this case, it is determined that an abnormal state has occurred, and an abnormal signal is output to the control means 4700.
Alternatively, a timer may be provided, and the person detection unit 250 may determine that an abnormality has occurred when a person is present in the immediate area of the safe for a predetermined time or longer (for example, 30 seconds or longer).

Next, a fourth embodiment according to the present invention will be described. In the first embodiment or the second embodiment, as shown in the schematic diagram of FIG. 7, a suspicious person who attempts to rob a gold item when the office is in a dark position and is in a straight line with the safe. Was supposed to be detected as.
A suspicious person having a different property can be exemplified by a person who illuminates the interior of an office room while the room lighting should normally be turned on and brightened. In this case, for example, it is considered that there is a preliminary intention of the crime such as searching for any valuable items other than the safe or for searching for important documents. In the fourth embodiment, such an office room is dark, and a person who is jealous of the inside of the office room is detected.

Since the schematic configuration diagram of the monitoring device according to the fourth embodiment can be realized in the same manner as in the first embodiment or the second embodiment, a description thereof will be omitted.
FIG. 11 shows a schematic diagram of the movement of a person to be detected in the fourth embodiment.

In FIG. 11, it is assumed that the position of the person at the previous time 750 is moved to the position 751 of the person at the current time. The locus is shown by a vector 752. A straight line obtained by extending the vector 752 is a dotted line indicated by reference numeral 753. In the first embodiment and the second embodiment, it is temporally determined that the closest region 6 exists at the point where the dotted line 753 is extended. In the fourth embodiment, conversely, as shown in FIG. 11, it is temporally determined that the closest region 6 does not exist at the point where the dotted line 753 is extended. It is determined whether or not the process continues. If it continues over time, it is determined that the person is wandering (roaming).

The processing flow is changed to the determination processing in step S750 in the flowchart shown in FIG. 6, and the approach counter 334 and the stay counter 335 are each 0, but the same person is continuously tracked over a predetermined time. In this case, the abnormality determination unit 46 may determine abnormality, and a detailed flowchart is omitted.
Alternatively, if it is desired to detect the state where the safe is in a straight line and determine that the state is abnormal, the values of the approach counter 334 and the stay counter 335 are also the same as in the first embodiment or the second embodiment. You may determine abnormality by referring to it.

The monitoring device according to the present invention is not limited to the embodiment described above.
As described above, the detection of a person in the office, which is the monitoring area, can be similarly realized by using an ultrasonic sensor or a laser light sensor in addition to the microwave sensor.
Alternatively, it is desirable to provide a plurality of transmitters / receivers, but it is also possible to acquire person position information by means of RFID.

Furthermore, instead of the non-contact type distance measuring sensor as described above, if circumstances allow, a mat-type sensor is spread all over the floor of the office so that the position that reacts to the person's weight is directly used as position information. It can also be used.
Moreover, even if there are a plurality of persons who perform suspicious behavior in the office, they can be processed in the same manner by using well-known methods in the image processing technology and sensor technology.

Furthermore, in the embodiment according to the present invention, the office room in the store is exemplified as the monitoring area. However, it is a normal operation to turn off the room lighting although it becomes unmanned in the released state and enters and exits freely. On the other hand, the monitoring apparatus of the present invention can be similarly applied to any place that can be regarded as suspicious when the room is entered with the room light turned off and a predetermined action is taken.
In addition, it is an operation with no distinction between security release and security set, and it can be applied to a place where an abnormal state is always detected while people can enter and leave freely.

Furthermore, both the act of approaching the safe, which is an important item, and the act of staying in the immediate area of the safe are used as conditions for reporting to the security center. A warning is given by the LED display that the approaching action is suspicious, and the monitoring device responds to the action by entering the immediate area of the safe in spite of the warning and notifying the security center etc. only after a certain period of time. You can also divide the levels.

DESCRIPTION OF SYMBOLS 1 ... Safe 2 ... Near field 10 ... Monitoring apparatus 20 ... Imaging part 30 ... Memory | storage part 40 ... Image signal processing part 45 ... Brightness determination means 50 ... Output Part

Claims (5)

A monitoring device for monitoring a monitoring object arranged in a monitoring area,
A brightness determination unit that detects the illuminance in the monitoring area and outputs the detected illuminance below a predetermined illuminance;
A person detection unit that outputs when the presence of a person is detected in the close area of the monitoring target;
An abnormality determination unit that determines that there is an output when there is an output from the person detection unit when the brightness determination unit outputs that the luminance is less than or equal to a predetermined illuminance;
A monitoring apparatus comprising: a control unit that outputs an abnormal signal when the abnormality determining unit determines that an abnormality has occurred. The monitoring apparatus according to claim 1, wherein the abnormality determination unit determines that the person is abnormal when the person is located in a close region of the monitoring target continuously for a predetermined time or more based on an output from the person detection unit. Furthermore, a storage unit that pre-stores the position of the monitoring object in the monitoring area;
A tracking unit for tracking a person existing in the monitoring area;
3. The abnormality determination unit according to claim 1 or 2, wherein the abnormality determination unit determines that the person is approaching the monitoring target object when the brightness determination unit outputs a predetermined illuminance or less. The monitoring apparatus according to any one of the above. A monitoring device for monitoring a monitoring object arranged in a monitoring area,
A brightness determination unit that detects the illuminance in the monitoring area and outputs the detected illuminance below a predetermined illuminance;
A person detection unit that outputs when the presence of a person is detected in the close area of the monitoring target;
A tracking unit for tracking a person existing in the monitoring area;
An abnormality determination unit that determines that the person is wandering in the monitoring area for a predetermined time when the brightness determination unit outputs that the luminance is less than or equal to a predetermined illuminance;
A monitoring apparatus comprising: a control unit that outputs an abnormal signal when the abnormality determining unit determines that an abnormality has occurred. The brightness determination unit determines illuminance from an average luminance of an image taken in the monitoring area,
The person detection unit extracts a person from the image and detects the presence of a person in the close region to be monitored from the extraction position;
The monitoring apparatus according to claim 3, wherein the tracking unit tracks a person from the images obtained by sequentially capturing monitoring areas.