JP5361014B2 - Traffic monitoring system - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To obtain a passage monitoring system capable of reliably monitoring the illegitimate invasion of a non-permitted person into an entry restriction area by eliminating the dead zone of a monitoring camera without using a plurality of monitoring cameras. <P>SOLUTION: A passage monitoring system is provided with: a camera 1 for taking an image of a region nearby a predetermined gateway from a first direction; a mirror 11 which is arranged in the field angle of the camera 1 so that a region nearby the gateway is reflected when viewed from a second direction different from the first direction; and a processing part 4 which determines whether the number of persons passing through the gateway is singular or plural on the basis of a first image reflected on the mirror 11 within an image taken by the camera 1 and a second image obtained by taking an image of the area nearby the gateway from the first direction with the camera 1. The processing part 4 determines whether the number of the persons is singular or plural on the basis of the first image using a first algorithm and determines whether the number of the persons is singular or plural on the basis of the second image using a second algorithm. <P>COPYRIGHT: (C)2013,JPO&amp;INPIT

Description

  The present invention relates to a traffic monitoring system, and more particularly to a traffic monitoring system for monitoring unauthorized entry into a restricted access area.

  2. Description of the Related Art Conventionally, in order to monitor unauthorized entry into a restricted access area, there is a system that captures a region near the entrance to the restricted access area with a monitoring camera and records the captured image on a hard disk or the like.

  The following Patent Documents 1 and 2 disclose an example of a monitoring system that performs room entry management using a camera.

JP 2006-209728 A JP 2006-243972 A

  In a surveillance system in which the area near the entrance to the entrance restricted area is photographed by a surveillance camera, the position of the occupant's face is the center of the field of view in order to clearly see and record the appearance (particularly the face) of the occupant. In many cases, surveillance cameras are installed in such a composition.

  FIG. 15 is a side view showing an installation example of the surveillance camera. In FIG. 15, the right side of the door 110 at the entrance is an access restriction area. The surveillance camera 101 is fixed to the upper wall of the door 110. In this example, the surveillance camera 101 is installed with a composition in which the optical axis center LC of the surveillance camera 101 faces obliquely downward.

  However, in such a monitoring system, a person who is allowed to enter the restricted access area in advance (hereinafter referred to as “permitted person”) is not permitted to enter the room (hereinafter referred to as “non-permitted person”). Even if a person who tries to infiltrate illegally (so-called “follow-up”), if a non-permitted person enters the blind spot of the surveillance camera generated by the permitter, the non-permitted person may be photographed by the surveillance camera. There is a problem that you can not.

  In addition, if the area near the entrance / exit is photographed from a plurality of directions by a plurality of monitoring cameras in order to eliminate the blind spot, the cost increases and communication control for synchronizing the plurality of monitoring cameras is necessary. There is a problem that processing becomes complicated.

  The present invention has been made to solve these problems. By eliminating the blind spots of the surveillance cameras without using a plurality of surveillance cameras, the unauthorized entry of unauthorized persons into the restricted access area is ensured. The purpose is to obtain a traffic monitoring system that can be monitored.

A traffic monitoring system according to a first aspect of the present invention is a camera that captures an area near a predetermined entrance / exit from a first direction, and a vicinity of the entrance / exit from a second direction different from the first direction within an angle of view of the camera. A mirror arranged so as to reflect an area; a first image projected on the mirror in an image captured by the camera; and a second image captured by the camera in the vicinity of the doorway from the first direction. And a processing unit that determines whether one or more persons pass through the doorway, and the first direction is a direction for photographing a person heading for the doorway from the front, and the second direction is a upper region near the doorway, the processing unit first performs the s judgment of the person using the first algorithm based on the first image, if the person in the judgment is determined to singular , Then, and performs s determination of a person using a second algorithm different from the first algorithm based on the second image.

  According to the traffic monitoring system according to the first aspect of the invention, the mirror is photographed so that the vicinity area of the entrance / exit is photographed by the camera from the first direction and the vicinity area of the entrance / exit is reflected from the second direction different from the photographing direction of the camera. Has been placed. Then, the processing unit determines whether the number of persons passing through the entrance is one or more based on the first video displayed on the mirror in the captured video. Therefore, the blind spot of the monitoring camera can be eliminated by using only one monitoring camera without using a plurality of monitoring cameras. As a result, unauthorized entry of unauthorized persons into the restricted access area can be reliably monitored.

  Further, according to the traffic monitoring system according to the first aspect, the processing unit displays the first image in which the image of the area near the entrance viewed from the second direction is reflected in the mirror and the area near the entrance in the first direction. It is determined whether one or more persons pass through the entrance / exit based on the second video imaged from the above. That is, it is possible to execute pass / fail determination of passers-by using a plurality of videos obtained by viewing the vicinity of the entrance / exit from a plurality of directions. As a result, it is possible to increase the accuracy of determination compared to the case where only one of the first video and the second video is used.

Further , according to the traffic monitoring system according to the first invention, since the mirror is arranged above the area near the entrance, the top of the passerby is projected in the first video. Therefore, regarding the first video, it is possible to determine whether there are one or more passers-by using an algorithm for specifying the top of the person. On the other hand, with respect to an image (second image) obtained by photographing the vicinity of the entrance / exit from the first direction by the camera, it is possible to determine whether there are one or more passers-by using an algorithm for specifying a person's face. . That is, according to the traffic monitoring system according to the first aspect of the present invention, it is possible to execute pass / fail determination of passers-by using different types of algorithms for the first video and the second video. As a result, the determination accuracy can be increased.

It is a front view showing the environment where the traffic monitoring system concerning an embodiment of the invention was introduced. It is a side view showing the environment where the traffic monitoring system concerning an embodiment of the invention was introduced. It is a block diagram which shows the structure of the signal processing part in the traffic monitoring system which concerns on embodiment of this invention. It is a side view which shows the condition where the nonpermitted person is going through the doorway after the permitted person. FIG. 5 is a view showing a part of image data picked up by a camera in correspondence with FIG. 4. 4 is a flowchart for explaining an operation of a processing unit shown in FIG. 3. It is a perspective view which shows the modification of the traffic monitoring system which concerns on embodiment of this invention. It is a figure which extracts and shows a part of image data imaged with the camera. It is a top view which shows the modification of the traffic monitoring system which concerns on embodiment of this invention. It is a side view which shows the modification of the traffic monitoring system which concerns on embodiment of this invention. It is a figure which shows an example of the image data extracted from the image data shown in FIG. It is a figure which shows the image data obtained by image | photographing a person from a 1st direction. FIG. 4 is a block diagram showing a modification of the configuration of the signal processing unit in correspondence with FIG. 3. It is a flowchart for demonstrating operation | movement of the process part shown in FIG. It is a side view which shows the example of installation of a surveillance camera.

  Hereinafter, embodiments of the present invention will be described in detail with reference to the drawings. In addition, the element which attached | subjected the same code | symbol in different drawing shall show the same or corresponding element.

  1 and 2 are a front view and a side view, respectively, showing an environment where a traffic monitoring system according to an embodiment of the present invention is introduced. The traffic monitoring system according to the present embodiment is a system for monitoring a person passing through an entrance / exit to an access restricted area (that is, an area where entry of persons other than those permitted to enter the room is prohibited). . Referring to FIGS. 1 and 2, a door 10 is provided at the entrance / exit of the restricted access area. In FIG. 1, the back side from the door 10 is an access restriction area. In FIG. 2, the right side of the door 10 is an access restriction area.

  On the wall above the door 10, a camera 1 capable of moving image shooting is installed. Referring to FIG. 2, camera 1 captures an area near the entrance / exit (specifically, an area near the entrance / exit in a non-restricted area that is not an entry restricted area) from the first direction indicated by arrow V <b> 1. . The mirror 11 is fixed at a predetermined angle at a predetermined location on the ceiling. The mirror 11 is arranged so as to be included in the shooting angle of view of the camera 1 (see the broken line in FIG. 2). When viewed from the position where the camera 1 is installed, the mirror 11 shows a view of the area near the entrance viewed from the second direction indicated by the arrow V2. Here, the second direction is upward (directly above or obliquely upward).

  FIG. 3 is a block diagram showing the configuration of the signal processing unit in the traffic monitoring system according to the present embodiment. An image processing unit 2 is connected to the rear stage of the camera 1. Further, an image recording unit 3 and a processing unit 4 are connected to the subsequent stage of the image processing unit 2. The image processing unit 2 and the processing unit 4 may be built in the camera 1. Alternatively, the image processing unit 2 and the processing unit 4 may be configured in the same apparatus as the image recording unit 3.

  An image (image data D <b> 1) in the vicinity of the entrance / exit captured by the camera 1 is input to the image processing unit 2. The image processing unit 2 performs general image processing such as gamma correction and white balance adjustment on the image data D1, and outputs image data D2. The image data D2 is input to the image recording unit 3 and recorded on an arbitrary recording medium such as a hard disk or a memory card.

  The image data D2 is input to the processing unit 4. The processing unit 4 includes an extraction unit 5 and a determination unit 6, and determines whether one or more persons pass through the entrance / exit based on the image data D <b> 2. Details of processing contents in the processing unit 4 will be described later.

  FIG. 4 is a side view showing a situation where the non-permitted person 21 is going through the doorway after the permitted person 20. Although not shown in FIG. 4, a card key is assigned to the permitter 20 in advance, and the card 10 is read by a card reader (not shown) installed in the vicinity of the door 10. Is unlocked so that the permitter 20 can pass through the doorway.

  FIG. 5 is a view showing a part of the image data D2 picked up by the camera 1 in correspondence with FIG. By photographing from the first direction indicated by the arrow V1, the face of the permitter 20 clearly appears in the image data D2. On the other hand, in the photographing from the first direction, the blind spot of the camera 1 generated by the permitter 20 becomes a blind spot, and the non-permitted person 21 does not appear clearly in the image data D2.

  2 and 4, since the mirror 11 is included in the shooting angle of view of the camera 1, the mirror 11 is included in the image data D3 which is a part of the image data D2 with reference to FIG. 11 is included. The mirror 11 in the image data D3 shows the state of the vicinity of the doorway as viewed from the second direction indicated by the arrow V2 in FIGS. Since the second direction is above the vicinity of the entrance / exit, as shown in FIG. 5, the non-permitted person 21 does not become the blind spot of the permitted person 20 but appears clearly in the image data D3.

  FIG. 6 is a flowchart for explaining the operation of the processing unit 4 shown in FIG. Hereinafter, the processing content in the processing unit 4 will be described in detail with reference to FIGS.

  First, in step SP11, the extraction unit 5 of the processing unit 4 illustrated in FIG. 3 inputs the image data D2 (see FIG. 5) from the image processing unit 2.

  Next, in step SP12, the extraction unit 5 extracts the image data D3 from the image data D2. The position of the image data D3 in the image data D2 is taught in advance to the extraction unit 5 by a method such as range designation using coordinates at the time of initial setting after the camera 1 and the mirror 11 are installed. The image data D3 extracted by the extraction unit 5 is input to the determination unit 6 of the processing unit 4.

  Next, in step SP13, the determination unit 6 determines whether one or more persons pass through the entrance / exit based on the image data D3. As an algorithm for the determination, any algorithm that can detect the number of persons by specifying the top of the person can be adopted. For example, a background image of the image data D3 in a state in which no person is reflected is captured in advance, and the difference between the background image and the image data D3 obtained by the current capturing is obtained. create. Next, after the binarization process is performed on the difference image, the luminance value is integrated for each row and each column of the difference image, thereby creating luminance distributions in the horizontal direction and the vertical direction of the difference image, respectively. Then, depending on whether the number of peaks in each luminance distribution is one or plural, it is determined whether the number of persons included in the image data D3 obtained by the current photographing is one or plural.

  Next, in step SP14, when there are a plurality of persons included in the image data D3, the determination unit 6 determines that there is an unauthorized intruder (non-permitted person 21) in addition to the permitted person 20. In this case, the process proceeds to step SP16, where a predetermined alarm is issued by an alarm device (not shown). On the other hand, when the number of persons included in the image data D3 is one, the determination unit 6 determines that there is no unauthorized intruder. In this case, the process proceeds to step SP15 and no warning is given by the warning device. In the above description, whether or not to issue an alarm is controlled based on the presence or absence of an unauthorized intruder, but the following control can also be performed. For example, when there is no unauthorized intruder, the image recording unit 3 records the image data D2 as a still image, and when there is an unauthorized intruder, it is recorded as a moving image. Alternatively, the image recording unit 3 records the image data D2 as a moving image. When an unauthorized intruder exists, the image recording unit 3 extends the recording time or records frames per unit time as compared to the case where there is no unauthorized intruder. Increase the number or decrease the compression rate of the image to improve the image quality.

  As described above, according to the traffic monitoring system according to the present embodiment, as shown in FIGS. 2 and 4, the area near the entrance / exit is photographed by the camera 1 from the first direction (arrow V <b> 1), and the camera 1 is photographed. The mirror 11 is arranged so that a region near the entrance / exit is reflected from a second direction (arrow V2) different from the direction. Then, the processing unit 4 shown in FIG. 3 determines whether one or a plurality of people pass through the doorway based on the video (image data D3) projected on the mirror 11 in the captured video (image data D2). . Therefore, the blind spot of the monitoring camera can be eliminated by using only one monitoring camera without using a plurality of monitoring cameras. As a result, unauthorized entry of unauthorized persons into the restricted access area can be reliably monitored.

<Modification 1>
FIG. 7 is a perspective view showing a modification of the traffic monitoring system according to the present embodiment. In the above description, the mirror 11 is installed above the vicinity of the doorway as shown in FIGS. 1 and 2, but instead of the mirror 11 (or in addition to the mirror 11) as shown in FIG. You may install the mirror 30 in the side of the vicinity area | region. The mirror 30 is arranged so as to be included in the shooting angle of view of the camera 1 (see the broken line in FIG. 7). When viewed from the position where the camera 1 is installed, the mirror 30 shows a view of the area near the entrance viewed from the side.

  FIG. 8 is a diagram showing a part of the image data D2 picked up by the camera 1 in a situation where the non-permitted person 21 is going through the doorway after the permitted person 20. As shown in FIG. 7, since the mirror 30 is included in the shooting angle of view of the camera 1, the mirror 30 is included in the image data D3 which is a part of the image data D2 with reference to FIG. ing. The mirror 30 in the image data D3 shows the state of the area near the entrance as viewed from the side. As shown in FIG. 8, the non-permitted person 21 does not become a blind spot of the permitted person 20 but clearly appears in the image data D3.

  The processing contents in the processing unit 4 are basically the same as described above. The extraction unit 5 of the processing unit 4 receives the image data D2 from the image processing unit 2 and extracts the image data D3 from the image data D2. The extracted image data D3 is input to the determination unit 6 of the processing unit 4, and the determination unit 6 determines whether one or more persons pass through the entrance / exit based on the image data D3. As an algorithm for determination, any algorithm that can detect the number of persons by specifying a person's face can be adopted. For example, a background image of the image data D3 in a state in which no person is reflected is captured in advance, and the difference between the background image and the image data D3 obtained by the current capturing is obtained. create. Next, after performing a skin color extraction process on the difference image to narrow down the face search range, a person's face is specified in the image data D3 by a template matching process using a person's profile profile. Then, the determination unit 6 determines whether the number of faces of the identified person is one or plural, and whether the number of persons included in the image data D3 obtained by the current photographing is one or plural. Determine whether. Since the subsequent operation is the same as described above, a description thereof is omitted here.

<Modification 2>
FIGS. 9 and 10 are a top view and a side view, respectively, showing a modification of the traffic monitoring system according to the present embodiment. In the above description, the camera 1 is installed on the wall above the door 10 as shown in FIGS. 1, 2 and 7, but the camera 1 is mounted on the main body of the card reader 40 as shown in FIGS. You may do it. As indicated by a broken line in FIG. 10, a mirror 11 is included in the shooting angle of view of the camera 1. Since the processing content in the processing unit 4 is the same as described above, description thereof is omitted here.

<Modification 3>
FIG. 11 is a diagram illustrating an example of the image data D3 extracted from the image data D2 illustrated in FIG. When the algorithm of specifying the head of a person by detecting a circle or an ellipse in the image is adopted, as shown in FIG. 11, in the situation where the two persons 50 and 51 are very close to each other, There is a possibility that the ellipse L including the two tops is detected as one top, and the person included in the image data D3 is erroneously detected as one person.

  FIG. 12 is a diagram illustrating image data D4 obtained by photographing the persons 50 and 51 from the first direction indicated by the arrow V1 in FIGS. That is, the image data D4 corresponds to a portion other than the image data D3 in the image data D2 shown in FIG. Referring to FIG. 12, the faces of the persons 50 and 51 appear in the image data D4 in a clearly distinguishable state. Therefore, it is possible to accurately detect that there are a plurality of persons included in the image data D4 by applying an algorithm for specifying a person's face to the image data D4.

  FIG. 13 is a block diagram showing a modification of the configuration of the signal processing unit corresponding to FIG. FIG. 14 is a flowchart for explaining the operation of the processing unit 4 shown in FIG. Hereinafter, the processing content in the processing unit 4 shown in FIG. 13 will be described in detail with reference to FIGS.

  First, in step SP21, the extraction unit 5 of the processing unit 4 inputs the image data D2 from the image processing unit 2.

  Next, in step SP22, the extraction unit 5 extracts image data D3 (see FIG. 11) and image data D4 (see FIG. 12) from the image data D2. The image data D3 and D4 extracted by the extraction unit 5 are input to the determination unit 6 of the processing unit 4.

  Next, in step SP23, the determination unit 6 determines whether one or more persons pass through the entrance / exit based on the image data D3. As an algorithm for the determination, any algorithm that can detect the number of persons by specifying the top of the person can be adopted as in the above embodiment.

  Next, in step SP24, the determination unit 6 determines that there is an unauthorized intruder when there are a plurality of persons included in the image data D3. In this case, the process proceeds to step SP28, where a predetermined alarm is issued by an alarm device (not shown).

  On the other hand, when it is determined that the number of persons included in the image data D3 is one, the process proceeds to step SP25, and the determination unit 6 determines whether one or more persons pass through the entrance / exit based on the image data D4. To do. As an algorithm for determination, any algorithm that can detect the number of persons by specifying a person's face can be adopted. For example, a background image of the image data D4 in a state in which no person is reflected is captured in advance, and the difference between the background image and the image data D4 obtained by the current capturing is obtained. create. Next, after performing a skin color extraction process on the difference image to narrow down the face search range, a person's face is specified in the image data D4 by a template matching process using a face template in front of the person. Then, the determination unit 6 determines whether the number of faces of the identified person is one or more, and the number of persons included in the image data D4 obtained by the current shooting is one or more. Determine whether.

  Next, in step SP26, the determination unit 6 determines that there is an unauthorized intruder when there are a plurality of persons included in the image data D4. In this case, the process proceeds to step SP28, where a predetermined alarm is issued by an alarm device (not shown).

  On the other hand, when there is only one person included in the image data D4, the determination unit 6 determines that there is no unauthorized intruder. In this case, the process proceeds to step SP27 and no alarm is given by the alarm device.

  According to the traffic monitoring system according to the third modified example, the processing unit 4 includes a video (image data D3) in which the video in the vicinity of the entrance viewed from the second direction (arrow V2) is projected on the mirror 11 and the entrance / exit. Based on the video (image data D4) taken in the first direction (arrow V1), it is determined whether one or more persons pass through the entrance. That is, the pass / fail determination of passers-by is performed using a plurality of videos in which a region near the doorway is viewed from a plurality of directions. As a result, the accuracy of determination can be improved as compared with the case where only one of the image data D3 and the image data D4 is used.

  In addition, with respect to the image data D3, a single pass / fail decision is made using an algorithm for specifying the top of the person's head, while an algorithm for specifying the person's face is used for the image data D4. The passer-by-passer judgment is made. That is, according to the traffic monitoring system according to Modification 3, it is possible to execute pass / fail determination of passers-by using different types of algorithms for image data D3 and image data D4. As a result, it is possible to increase the accuracy of determination compared to the case where the same type of algorithm is used.

  Here, also in the said modification 3, the mirror 30 reflected from the side can be installed instead of the mirror 11 which images the area | region of the entrance / exit vicinity from upper direction. In this case, as with the image data D4, the pass / fail determination of passers-by is also performed for the image data D3 using an algorithm for specifying a person's face. In this way, by using the same kind of algorithm for the image data D3 and the image data D4 to perform pass / by determination of passers-by, the algorithms can be unified and the processing can be simplified.

1 Camera 4 Processing Unit 5 Extraction Unit 6 Determination Unit 10 Door 11, 30 Mirror

Claims (1)

A camera that captures a region near a predetermined doorway from the first direction;
A mirror disposed within the angle of view of the camera so that a region near the doorway is reflected from a second direction different from the first direction;
A person who passes through the entrance based on a first image projected on the mirror in an image photographed by the camera and a second image obtained by photographing the area near the entrance from the first direction by the camera. And a processing unit for determining whether or not there is one or more,
The first direction is a direction for photographing a person heading for the entrance from the front,
The second direction is above a region near the doorway;
The processing unit first performs single determination of a person using a first algorithm based on the first video, and if it is determined that the person is singular in the determination, next, based on the second video A traffic monitoring system that performs single / multiple determination of a person using a second algorithm different from the first algorithm.

Images