JP4622301B2 - Surveillance system and surveillance camera - Google Patents

Description

  The present invention relates to a monitoring system that ensures the security of facilities such as airports, banks, office buildings, condominiums, and the like where many people come and go, and a monitoring camera applied to the monitoring system.

  Conventionally, in facilities such as airports, banks, office buildings, and condominiums where a large number of people come and go, a monitoring system that monitors the entry and exit of suspicious persons using captured images taken by a surveillance camera is used to ensure security. We are using. In the conventional monitoring system, the entrance / exit of the facility, the main location in the facility, and the like are set as the monitoring area, and a monitoring camera is provided for each monitoring area. In addition, a host device is provided in a security room or the like, and the host device is configured so that captured images of each monitoring area can be confirmed. Specifically, the monitoring camera is connected to the host device via a LAN, an IP network, or the like. The monitoring camera transmits the captured image of the monitoring area being imaged to the host device in substantially real time. The captured image transmitted from the monitoring camera to the host device is a moving image. The host device stores the captured image (moving image) of the monitoring area transmitted for each monitoring camera.

  If a person is captured in the captured image of the monitoring area transmitted from the surveillance camera, the person's face image is cut out and collated with a pre-stored registrant's face image to be a registrant. A monitoring system is also put into practical use in which a host device is provided with a face collation function for determining whether or not to increase security. The registrant is set according to the use of the monitoring system. For example, if the facility that uses the surveillance system is an office building or apartment, etc., the person who is allowed to enter and exit the facility may be a registered person, and if it is an airport or bank, etc. A registered person (criminal) may be used. The host device outputs the result of face matching. For example, if the person whose face has been collated is a person who has not been permitted to enter or leave the facility, or if the person is a criminal who has been nominated, a warning or the like is issued to notify the security guard. Thereby, a security guard etc. can respond quickly to an unbelieving person, and can raise security.

Further, in the conventional monitoring system, it is detected whether or not a person is captured in the captured image. If the person is captured, the captured image is recorded on a recording medium (see Patent Document 1), or a background image. By compressing and storing the difference image between the reference frame and the captured image (see Patent Document 2), there is no omission of recording of the human image captured by the surveillance camera, and it is necessary to record the captured image. It has been proposed to reduce the recording capacity of various recording media.
JP 2000-115389 A JP 2000-13742 A

  However, in a monitoring system in which a large number of monitoring cameras are connected to a host device that is a host device, each monitoring camera transmits a captured image of the monitoring area in real time to the host device. The network traffic connecting the For this reason, there is a high possibility that the captured image of the monitoring area transmitted from the monitoring camera to the host device will be delayed or lost. When a delay or disappearance occurs in the captured image of the monitoring area transmitted from the monitoring camera to the host device, the host device cannot properly monitor the monitoring area, and there is a problem that security is lowered.

  On the other hand, the captured image of the monitoring area transmitted from the monitoring camera to the host device may be a rough image, or further compressed and transmitted to suppress the increase in traffic and solve the above problem. In this case, since the captured image of the monitoring area transmitted from the monitoring camera to the host device is rough, the face image of the person cut out from this captured image cannot be properly matched, and as a result, it is possible to prevent a decrease in security. Can not.

  The object of the present invention is applicable to a monitoring system capable of appropriately monitoring a monitoring area in a host device and preventing a decrease in security even when there are many surveillance cameras connected to the host device, and to this monitoring system. It is to provide a surveillance camera.

  In order to solve the above problems, the monitoring system of the present invention has the following configuration.

(1) In a monitoring system in which a plurality of monitoring cameras are connected to a host device via a network ,
The surveillance camera is
Imaging means for imaging the surveillance area;
A moving image storage unit that stores a captured image of the monitoring area captured by the imaging unit as a moving image with a time stamp indicating an imaging date and time ;
If a person is imaged in the captured image of the monitoring area imaged by the imaging means, face image extraction means for extracting the person's face image;
Time stamp adding means for adding a time stamp indicating the imaging date and time of the captured image obtained by extracting the face image to the face image extracted by the face image extracting means;
The face image cut-out unit cutout, and a still picture transmission means for transmitting to said host system in a still image a face image in which the time stamp adding unit has added a time stamp,
The host device is
For a face image transmitted as a still image from the surveillance camera, when a moving image acquisition request for this face image is input, transmission of the moving image of the date and time indicated by the time stamp added to the face image is performed. Provided with a video request means for requesting a surveillance camera;
Furthermore, the surveillance camera is
Upon receiving the video transmission request from the previous SL upper apparatus, the predetermined time of moving across the time indicated by the time stamp included in the received transmission request, read out from said video storage means, to said upper device A moving image transmitting means for transmitting is provided.

In this configuration, the surveillance camera stores the captured image of the monitoring area captured by the imaging unit as a moving image in the moving image storage unit. In addition, if a person is captured in the captured image of the monitoring area by the face image cutout unit, the surveillance camera cuts out the face image of the person and freezes the cutout face image to the host device by the still image transmission unit. Send as a picture. A time stamp indicating the imaging date and time is added to the moving image and the still image.

The host device can confirm the date and time when the image cut out from the face image is captured by the time stamp added to the face image transmitted as a still image from the surveillance camera. In addition, when a moving image acquisition request for a face image transmitted from the monitoring camera as a still image is input to the host device , the host device requests the monitoring camera to transmit the moving image for the face image. This request includes the date and time indicated by the time stamp added to the face image. When the monitoring camera receives the moving image transmission request from the host device, the monitoring camera reads out the moving image for a predetermined time with the date and time indicated by the time stamp included in the received transmission request from the moving image storage means, and the host device Send to. For example, the predetermined time may be a 4 to 5 s moving image centered on the time indicated by the time stamp.

  As described above, when a person is captured in the captured image of the monitoring area, the monitoring camera cuts out the face image of the person in substantially real time, and transmits the cut-out face image as a still image to the host device. Therefore, the traffic on the network connecting the monitoring camera and the host device can be sufficiently suppressed without roughening the face image transmitted from the monitoring camera to the host device. Therefore, it is possible to prevent a situation where the monitoring area cannot be properly monitored in the host device, and security is not lowered.

In addition, the host device can acquire a moving image related to a face image transmitted as a still image from the monitoring camera as necessary, so that the situation when the person is passing through the monitoring area is recorded as a moving image. Can be confirmed.
As for the face image cutout means of the monitoring camera, if a person is captured in the captured image of the monitoring area captured by the imaging means, a plurality of frames are stored in the buffer of the captured image in which the person is captured The captured image from which the face image is cut out may be selected from the captured images of a plurality of frames stored in the buffer. In this way, it is possible to prevent the face image of the person being cut out from the captured image in which the face of the person being picked up is cut off or the face orientation is not appropriate.
Furthermore, the moving image storage means of the monitoring camera may be configured to store the captured images of a plurality of frames stored in the buffer as moving images in the monitoring area. In this way, the storage capacity of the moving image storage means can be suppressed.

(2) The host device includes a face image storage means for storing a registrant's face image;
A face collation unit that collates a face image transmitted as a still image from the monitoring camera with a registrant's face image stored in the face image storage unit and outputs the result. .

  In this configuration, the host device collates the face image of the person transmitted as a still image from the surveillance camera with the face image of the registrant and outputs the result, so when a suspicious person or the like is detected Therefore, it is possible to promptly notify the security guard or the like, and it is possible to promptly respond to the security guard or the like.

(3) The host device includes a face image storage unit that stores a face image transmitted as a still image from the monitoring camera;
Display means for displaying the face image stored in the face image storage means,
The moving image requesting unit receives an input relating to a moving image acquisition request for the face image displayed by the display unit.

  In this configuration, in the host device, the face image transmitted as a still image from the monitoring camera can be stored in the face image storage means, and the stored face image can be displayed on the display means. For the face image displayed by the display unit, the moving image request unit accepts an input related to a moving image acquisition request. As a result, the confirmation of the face image of a person sent as a still image from the surveillance camera and the confirmation of the situation when this person has passed the surveillance area are not only the timing when the face image is sent from the surveillance camera. You can do it whenever you want.

( 4 ) In a surveillance system in which a plurality of surveillance cameras are connected to a host device via a network ,
The surveillance camera is
Imaging means for imaging the surveillance area;
A moving picture transmitting means for sending a picked-up image of the monitoring area picked up by the image pick-up means to the high-order apparatus as a moving picture to which a time stamp indicating the image pickup date and time is added ;
If a person is imaged in the captured image of the monitoring area imaged by the imaging means, face image extraction means for extracting the person's face image;
Time stamp adding means for adding a time stamp indicating the imaging date and time of the captured image obtained by extracting the face image to the face image extracted by the face image extracting means;
Still image storage means for storing the face image with the face image cutout means cut out and the time stamp adding means with the time stamp added as a still image;
The moving image transmitting means transmits the moving image as an image coarser than the face image stored in the still image storage means,
In addition, the host device is
When a face image acquisition request is input for a person imaged in the moving image transmitted from the surveillance camera, the date and time indicated by the time stamp added to the position where the person is imaged in the moving image is input . Still image request means for requesting the surveillance camera to transmit a face image,
Furthermore, the surveillance camera is
Upon receiving the transmission request of the face image from the previous SL upper apparatus, a face image closest timestamp date indicated by the time stamp included in the received transmission request is added, from the still image memory means Still image transmission means for reading and transmitting to the higher-level device is provided.

In this configuration, the surveillance camera transmits the captured image of the monitoring area captured by the imaging unit to the host device by the moving image transmission unit as a coarse image. In addition, if a person is captured in the captured image of the monitoring area by the face image cutout unit, the surveillance camera cuts out the face image of the person and stores the cut out face image in the still image storage unit. A time stamp indicating the shooting date and time is added to the moving image and the still image (face image).

  On the other hand, when a request for acquiring a face image is input for a person captured in a moving image transmitted from the monitoring camera, the host device requests the monitoring camera to transmit the face image. The surveillance camera transmits the face image stored in the still image storage unit as a still image in response to a request from the host device.

  As described above, since the surveillance camera is configured to transmit the captured image of the surveillance area to the host device in a substantially real time, the traffic on the network connecting the surveillance camera and the host device can be sufficiently suppressed. it can. In addition, the moving image transmitted from the monitoring camera to the host device is not only a coarse image but also compressed, so that traffic on the network connecting the monitoring camera and the host device can be further suppressed. In the host device, the face image (still image) of the person captured in the moving image transmitted from the monitoring camera can be acquired from the monitoring camera as necessary, and the face image is not a rough image such as a moving image. Therefore, by acquiring this face image, face collation for this person can be performed properly. Therefore, it is possible to prevent a situation where the monitoring area cannot be properly monitored in the host device, and security is not lowered.

  According to the present invention, it is possible to sufficiently suppress traffic on the network connecting the monitoring camera and the host device. Therefore, it is possible to prevent a situation where the monitoring area cannot be properly monitored in the host device, and security is not lowered.

  Hereinafter, a monitoring system according to an embodiment of the present invention will be described.

  FIG. 1 is a diagram showing a configuration of a monitoring system according to an embodiment of the present invention. In the monitoring system of this embodiment, a plurality of monitoring cameras 1 are connected to a host device 2 via a network 3. This monitoring system is used to ensure security by monitoring persons entering and leaving facilities such as airports, banks, office buildings, and condominiums, and persons passing through the facilities. The surveillance camera 1 is provided for each surveillance area. The monitoring area is the entrance / exit of the facility or the main location within the facility. The host device 2 is installed in a security room or the like, and is configured such that a security guard or the like can check the status of the monitoring area. The network 3 is, for example, a LAN or an IP network.

  FIG. 2 is a diagram showing a configuration of a monitoring camera applied to the monitoring system according to the embodiment of the present invention. The monitoring camera 1 determines whether or not a person is captured in the captured image of the monitoring area captured by the imaging unit 12, the control unit 11 that controls the operation of the main body, the imaging unit 12 that images the monitoring area, and the person If the image is captured, a Face Finding module 13 (hereinafter referred to as FF module 13) that cuts out the face image of the captured person as a still image, and a captured image of the monitoring area captured by the imaging unit 12 as a moving image An image compression unit 14 that acquires and compresses the compressed image, a moving image storage unit 15 that stores a moving image compressed by the image compression unit 14, and a communication unit 16 that communicates with the host device 2 via the network 3. It has.

  The imaging unit 12 has a CCD as an image sensor. The captured image of the monitoring area captured by the imaging unit 12 is input to the FF module 13 and the image compression unit 14. The FF module 13 and the image compression unit 14 capture captured images input at a predetermined time interval, for example, 0.3 s interval. The time intervals at which the FF module 13 and the image compression unit 14 capture captured images may be the same or different. The FF module 13 has a function of determining whether or not a person is captured in the captured image, and a function of cutting out a face image of the person captured when the person is captured. The FF module 13 corresponds to the face image cutting means referred to in the present invention. The face image (still image) clipped from the captured image by the FF module 13 is transmitted from the communication unit 16 to the host device 2 via the network 3 in substantially real time. The face image transmitted to the host device 2 is added with a time stamp indicating the imaging date and time. Details of this time stamp will be described later. The image compression unit 14 compresses the captured image of the monitoring area acquired as a moving image by MPEG4 or the like. The moving image storage unit 15 stores the compressed moving image of the monitoring area. The moving image stored in the moving image storage unit 15 is also added with a time stamp indicating the date and time when the image was captured at an appropriate interval, and the shooting date and time of the moving image stored in the moving image storage unit 15 can be determined. In response to a request from the host device 2, the monitoring camera 1 transmits the moving image stored in the moving image storage unit 15 to the host device 2. The moving image storage unit 15 has, for example, a hard disk as a recording medium.

  Next, the configuration of the host device 2 applied to this monitoring system will be described. FIG. 3 is a diagram showing the configuration of the host device applied to the monitoring system according to the embodiment of the present invention. The host device 2 includes a control unit 21 that controls the operation of the main body, a registrant data storage unit 22 that stores a face image for each registrant, a communication unit 23 that communicates with the monitoring camera 1 via the network 3, and a monitoring unit. A face collation unit 24 that collates a face image transmitted as a still image from the camera 1 with a registrant's face image stored in the registrant data storage unit 22, and MPEG 4 transmitted from the surveillance camera 1. An image decompression unit 25 that decompresses a compressed video, a display unit 26 that displays a face image or video transmitted from the surveillance camera 1, and a face that stores a face image transmitted as a still image from the surveillance camera 1 An image storage unit 27.

  The registrant data storage unit 22 stores feature quantities of facial parts such as eyes, nose, and mouth as registrant face images. The registrant is, for example, a person who is permitted to enter or leave the facility, or is a criminal who has been appointed, and determines the registrant according to the use of the monitoring system. The face matching unit 24 extracts the feature amount of the face part from the face image transmitted as a still image from the monitoring camera 1, and the feature amount of each registrant's face part stored in the registrant data storage unit 22. The degree of similarity is calculated, and it is determined that the registrant who has the maximum degree of similarity and is larger than a predetermined threshold is the person who has transmitted the face image this time. If there is no registrant whose similarity is greater than a predetermined threshold, it is determined that the user is not a registrant. The image decompression unit 25 decompresses the moving image transmitted from the monitoring camera 1 and generates a reproduction image signal. The display unit 26 displays a face image or a moving image (an image based on a reproduction image signal generated by the image expansion unit 25) transmitted from the monitoring camera 1. The face image storage unit 27 stores a face image transmitted as a still image from the monitoring camera 1. The face image storage unit 27 has a recording medium such as a hard disk.

  Hereinafter, the operation of the monitoring system of this embodiment will be described.

  The monitoring camera 1 performs in parallel a monitoring image process for processing a captured image in a monitoring area in substantially real time and a moving image transmission process for transmitting a moving image stored in the moving image storage unit 15 in response to a request from the host device 2. To do. First, monitoring image processing in the monitoring camera 1 will be described. The monitoring image processing includes moving image processing for storing a captured image of the monitoring area captured by the imaging unit 12 in the moving image storage unit 15 as a moving image, and if a person is captured in the captured image, the face image of the person is displayed. These are two face image transmission processes that are cut out and transmitted to the host device 2.

  FIG. 4 is a flowchart showing the moving image processing. The monitoring camera 1 captures captured images of the monitoring area captured by the imaging unit 12 into the image compression unit 14 at predetermined time intervals, for example, at intervals of 0.3 s (s1, s2). The image compressing unit 14 converts the captured image captured at regular time intervals into a moving image, and compresses it with MPEG4 or the like (s3). The surveillance camera 1 stores the compressed moving image in the moving image storage unit 15 (s4). Time stamps indicating the shooting date and time are added to the moving images stored in the moving image storage unit 15 at appropriate intervals, for example, at intervals of 1 to 2 minutes.

  The monitoring camera device 1 stores the moving image of the monitoring area in the moving image storage unit 15 by repeatedly executing the processes of s1 to s4. In addition, since the time stamp indicating the shooting date and time is added to the moving image stored in the moving image storage unit 15 at an appropriate interval, the moving image shooting date and time stored in the moving image storage unit 15 is referred to by referring to this time stamp. Can be judged.

  Next, the face image transmission process will be described. FIG. 5 is a flowchart showing the face image transmission process. The monitoring camera 1 captures captured images of the monitoring area captured by the imaging unit 12 into the FF module 13 at predetermined time intervals, for example, at intervals of 0.3 s (s11, s12). The FF module 13 calculates the difference between the captured image captured last time and the captured image captured this time (s13). As is well known, when a person is imaged, the difference between the captured image captured last time and the captured image captured this time changes because the position and orientation of the person being imaged have changed. Become.

  If the difference calculated in s13 is larger than a predetermined threshold (s14), the FF module 13 determines that a person is captured in the captured image, and temporarily stores the captured image in a buffer memory (not shown). (S15). If the difference calculated in s13 is not larger than a predetermined threshold, the FF module 13 jumps to s19 and clears the buffer memory. When the captured image is stored in the buffer memory in s15, the FF module 13 determines whether or not the captured image having the appropriate number of frames for clipping the face image is stored in the buffer memory (s16). In s16, a captured image in which a face image appropriate for use in the face collation of the person cannot be cut out, for example, a captured image of a person whose face is cut or whose face orientation is not appropriate is displayed. In order not to cut out the face image, the captured image of several frames, for example, 4 to 6 frames, is stored in the buffer memory. In other words, the possibility that a captured image that can cut out a face image appropriate for use in face collation of the person is stored in the buffer memory is increased. If the FF module 13 determines in step s16 that the captured image having the appropriate number of frames is not stored in the buffer memory, the FF module 13 returns to step s11 and repeats the above processing to cut out the face image in this buffer memory. The captured image having an appropriate number of frames is stored.

  Note that the appropriate number of frames for extracting the face image is equal to or less than the number of frames of the captured image that can be captured by the FF module 13 within the time (2 to 3 s) during which the person being imaged passes through the monitoring area.

  On the other hand, if it is determined in s16 that the captured image having the appropriate number of frames for clipping the face image is stored in the buffer memory, the face image is extracted from the plurality of frames of captured images stored in the buffer memory. A captured image that is optimal for extraction is selected (s17). In s17, a captured image in which the face of the person being imaged is not cut and the face direction is the front is selected. The FF module 13 cuts out the face image of the person being captured from the captured image selected in s17 (s18). In s18, a face shape (elliptical shape) is detected from the captured image, and a range including the detected face shape is cut out as a face image of the person captured this time. When the FF module 13 cuts out the face image of the person imaged this time in s18, the FF module 13 clears the buffer memory (s19). The surveillance camera 1 adds a time stamp indicating the imaging date and time to the face image cut out in s18 (s20), transmits it to the host device 2 from the communication unit 16 (s21), returns to s11, and repeats the above processing.

  Thereby, the monitoring camera 1 can transmit the face image of a person passing through the monitoring area imaged by the imaging unit 12 to the host device 2 in substantially real time. In other words, the host device 2 can acquire a face image of a person passing through the monitoring area in substantially real time. Further, since the monitoring camera 1 does not transmit a moving image of the monitoring area to the host device 2 but transmits the cut face image as a still image, a sufficient amount of data is transmitted from the monitoring camera 1 to the host device 2. Even in a monitoring system in which there are many monitoring cameras 1 connected to the host device 2, traffic on the network 3 connecting the host device 2 and the monitoring camera 1 can be sufficiently suppressed. In addition, since the face image transmitted from the monitoring camera 1 to the host device 2 as a still image does not have to be a rough image, the face of the person imaged in the host device 2 to be described later can be appropriately matched and the security is lowered. I will not let you.

  Here, the operation of the host apparatus 2 that has received a face image transmitted as a still image from the monitoring camera 1 will be described. FIG. 6 is a flowchart showing face image reception processing in the host device. Upon receiving the face image transmitted as a still image from the monitoring camera 1, the host device 2 stores the face image in the face image storage unit 27 and displays the face image on the display unit 26 (s31, s32). ). The face image stored in the face image storage unit 27 is added with a time stamp indicating the imaging date and time of the captured image obtained by cutting out the face image. The host device 2 performs face collation processing for collating the face image transmitted from the surveillance camera 1 this time with the registrant's face image stored in the registrant data storage unit 22 (s33). This face matching process is performed using, for example, a well-known Gabor Wavelet and a Graph Matching Method. Here, although detailed description is omitted, in brief, feature amounts of facial parts such as eyes, nose and mouth are extracted from the face image transmitted from the monitoring camera 1, and the registrant data storage unit 22 is extracted. The degree of similarity with the feature amount of the facial part is calculated for each registrant stored in. Then, a registrant whose calculated similarity is maximum and is larger than a predetermined threshold is determined to be a person who has transmitted the face image this time. If there is no registrant whose similarity is greater than a predetermined threshold, it is determined that the user is not a registrant.

  The host device 2 outputs the result of the face collation process relating to s33 (s34). In s34, the result of face collation is displayed on the display unit 26, or an alarm is given if necessary. For example, if the facility prohibits persons other than the registrant from entering and exiting, a warning is issued when it is verified in s33 that the person who sent the face image this time is authenticated as a registrant. As described above, since the monitoring camera 1 transmits a face image of a person passing through the monitoring area in substantially real time to the host device 2, face matching for the person can be performed in substantially real time. Therefore, it is possible to detect in real time that a suspicious person has entered a facility, and to promptly notify a security guard or the like. Thereby, a security guard etc. can respond to a suspicious person promptly, and does not reduce the security of a facility.

  In addition, the host device 2 can cause the display unit 26 to display a face image transmitted as a still image from the monitoring camera 1 and stored in the face image storage unit 27. With respect to the person whose face image is displayed on the display unit 27, a moving image when the person is passing through the monitoring area can be acquired from the monitoring camera 1. A moving image acquisition process in which the host device 2 acquires a moving image from the monitoring camera 1 will be described. FIG. 7 is a flowchart showing the moving image acquisition process in the host device, and FIG. 8 is a flowchart showing the moving image transmission process of the monitoring camera corresponding to this.

  A security guard or the like inputs a moving image acquisition request to the host apparatus 2 for a person whose face image is determined to be necessary to confirm the situation when the user is passing through the monitoring area. Specifically, a face image of a person who acquires a moving image is displayed on the display unit 26 of the host device 2, and an input operation for requesting acquisition of the moving image is performed. When this moving image acquisition request is input, the host device 2 starts the moving image acquisition process shown in FIG. First, the time stamp added to the face image of the person who is requested to acquire the moving image is read (s41). The host device 1 transmits a transmission request for the moving image having the date and time indicated by the read type stamp in s41 from the communication unit 23 to the monitoring camera 1 (s42).

  When the monitoring camera 1 receives the moving image transmission request from the host device 2 (s51), it reads the date and time indicated by the time stamp included in the transmission request (s52). The surveillance camera 1 reads from the moving image storage unit 15 a moving image of several seconds before and after the date and time read in s52, for example, 4 to 5 seconds, for example, 4 to 5 seconds from the moving image storage unit 15 (s53). (S54).

  When the host device 2 receives the moving image transmitted from the monitoring camera 1 (s43), the host device 2 expands the moving image received by the image expansion unit 25 and generates a reproduced video signal (s44). The host device 2 displays an image based on the reproduced video signal generated in s44 on the display unit 26 (s45).

  As described above, the host device 2 can acquire and check a moving image of the person whose face image has been transmitted as a still image as it passes through the monitoring area as necessary. In other words, for the person whose face image is transmitted as a still image, the host device 2 can check the situation when the person is passing through the monitoring area. In addition, since the moving image that the monitoring camera 1 transmits to the host device 2 in response to this moving image transmission request is a moving image of several seconds, the traffic on the network 3 that connects the monitoring camera 1 and the host device increases unnecessarily. It is suppressed enough.

In the above embodiment, although the storing constantly moving of the monitor area in the moving picture storage unit 15, in the face image transmission processing, was cut out face image in s 18, before clearing the buffer memory in s 19 The captured images of a plurality of frames stored in the buffer memory may be compressed as a moving image and stored in the moving image storage unit 15. In other words, a moving image when a person is captured in the captured image may be stored in the moving image storage unit 15, and a moving image in a monitoring area when a person is not captured may not be stored in the moving image storage unit 15. In this way, the storage capacity required for the moving image storage unit 15 can be reduced. Further, since the moving image processing shown in FIG. 4 need not always be performed, the load on the monitoring camera 1 can be reduced.

  Next, another embodiment of the present invention will be described. The monitoring system of this embodiment has the configuration shown in FIG. 1 as in the above embodiment. The surveillance camera device 1 has substantially the same configuration as the device of the above embodiment, but differs in that it does not include the moving image storage unit 15 and a still image storage unit 17 as shown in FIG. Yes. The still image storage unit 17 stores the face image of the person cut out by the FF module 13. The host device 2 has substantially the same configuration as the device of the above embodiment, but differs in that it does not include the face image storage unit 27 and a moving image storage unit 28 as shown in FIG. ing. The moving image storage unit 28 stores the moving image of the monitoring area transmitted from the monitoring camera 1.

  Hereinafter, the operation of the monitoring system of this embodiment will be described.

  The monitoring camera 1 performs monitoring image processing for processing a captured image in a monitoring area in substantially real time, and a person's face image stored in the still image storage unit 17 in response to a request from the host device 2 as a still image. The still image transmission processing to be transmitted to 2 is executed in parallel. First, monitoring image processing in the monitoring camera 1 will be described. In this monitoring image processing, moving image transmission processing for transmitting the captured image of the monitoring area captured by the imaging unit 12 to the host device 2 as a moving image, and if a person is captured in the captured image, the person's face image Is stored and stored in the still image storage unit 17.

  FIG. 11 is a flowchart showing the moving image transmission process. The monitoring camera 1 captures captured images of the monitoring area captured by the imaging unit 12 into the image compression unit 14 at predetermined time intervals, for example, at intervals of 0.3 s (s61, s62). The image compression unit 14 converts the captured image captured at regular time intervals into a moving image and compresses it with MPEG4 or the like (s63). In s63, the captured image is compressed into a relatively coarse image. The monitoring camera 1 transmits the moving image compressed by the image compression unit 14 from the communication unit 16 to the host device 2 via the network 3 (s64). A time stamp indicating the imaging date and time is added to the moving image transmitted to the host device 2.

  The monitoring camera device 1 transmits the moving image of the monitoring area to the host device 2 in substantially real time by repeatedly executing the processes of s1 to s4. Further, since the moving image to be transmitted to the host device 2 is a rough image, the data amount of the moving image to be transmitted to the host device 2 can be sufficiently suppressed. Therefore, even in a monitoring system in which many monitoring cameras 1 are connected to the host device 2, traffic on the network 3 that connects the host device 2 and the monitoring camera 1 is sufficiently suppressed. In addition, since a time stamp indicating the imaging date and time is added to the moving image to be transmitted to the host device 2, the host device 2 can determine the imaging date and time by referring to the time stamp.

  Next, the face image storage process will be described. FIG. 12 is a flowchart showing the face image storage process. The monitoring camera 1 captures captured images of the monitoring area captured by the imaging unit 12 into the FF module 13 at predetermined time intervals, for example, at intervals of 0.3 s (s71, s72). The FF module 13 calculates the difference between the captured image captured last time and the captured image captured this time (s73). If the difference calculated in s73 is larger than a predetermined threshold (s74), the FF module 13 determines that a person is captured in the captured image, and temporarily stores the captured image in a buffer memory (not shown). (S75). If the difference calculated in s73 is not greater than a predetermined threshold value, the FF module 13 jumps to s79 and clears the buffer memory. When the captured image is stored in the buffer memory in s75, the FF module 13 determines whether the captured image having an appropriate number of frames for clipping the face image is stored in the buffer memory (s76). In s76, a captured image in which a face image appropriate for use in the face collation of the person cannot be cut out, for example, a captured image of a person whose face is cut or whose face orientation is not appropriate, is displayed. In order not to cut out the face image, the captured image of several frames, for example, 4 to 6 frames, is stored in the buffer memory. In other words, the possibility that a captured image that can cut out a face image appropriate for use in face collation of the person is stored in the buffer memory is increased. If the FF module 13 determines in step s76 that the captured image having the appropriate number of frames for cutting out the face image is not stored in the buffer memory, the process returns to step s71 and repeats the above processing to cut out the face image in this buffer memory. The captured image having an appropriate number of frames is stored.

  Note that the appropriate number of frames for extracting the face image is equal to or less than the number of frames of the captured image that can be captured by the FF module 13 within the time (2 to 3 s) during which the person being imaged passes through the monitoring area.

  On the other hand, if it is determined in s76 that the captured image having the appropriate number of frames for clipping the face image is stored in the buffer memory, the face image is extracted from the plurality of frames of captured images stored in the buffer memory. A captured image that is optimal for extraction is selected (s77). In s77, a captured image in which the face of the person being imaged is not cut and the face direction is the front is selected. The FF module 13 cuts out the face image of the person being captured from the captured image selected in s77 (s78). In s78, a face shape (elliptical shape) is detected from the captured image, and a range including the detected face shape is cut out as a face image of the person imaged this time. Further, the face image is cut out without making the captured image rough. In other words, the image compression unit 14 transmits to the host device 2 a moving image that has been compressed from a captured image to a rough image, but the FF module 13 cuts out a face image from the captured image without making the captured image a rough image. The FF module 13 clears the buffer memory when the face image of the person imaged this time is cut out in s78 (s79). The surveillance camera 1 adds a time stamp indicating the imaging date and time to the face image cut out in s78 (s80), stores it in the still image storage unit 17 (s81), returns to s71, and repeats the above processing.

  Thereby, the monitoring camera 1 can store the face image of a person passing through the monitoring area imaged by the imaging unit 12 in the still image storage unit 17 as a still image. Further, this face image is a non-rough image that can perform face matching properly.

  Here, the operation of the host apparatus 2 that has received the moving image transmitted from the monitoring camera 1 will be described. FIG. 13 is a flowchart showing the moving image reception process in the host device. When the host device 2 receives the moving image transmitted from the monitoring camera 1, the host device 2 stores the moving image in the moving image storage unit 28 (s91), and expands the received moving image in the image expansion unit 25 to generate a reproduced video signal. (S92). The host device 2 displays the moving image based on the reproduced video signal generated in s92 on the display unit 26 (s93).

  Thus, since the moving image of the monitoring area is displayed in the host device 2 in substantially real time, a guard or the like can check the status of the monitoring area in the host device 2 in substantially real time. Further, when the security guard determines that the face matching is necessary for the person of the moving image displayed on the display unit 26, the guard can acquire a face image that can appropriately perform the face matching for the person. Hereinafter, a face image acquisition process for acquiring a face image that can be appropriately face-matched in the host device 2 will be described.

  FIG. 14 is a flowchart showing face image acquisition processing in the host device, and FIG. 15 is a flowchart showing face image transmission processing of the monitoring camera corresponding thereto.

  A security guard or the like inputs a face image acquisition request to the host device 2 when determining that face matching is necessary for a person passing through the monitoring area. Specifically, an input operation for requesting acquisition of a face image is performed for a person displayed as a moving image on the display unit 26 of the host device 2. When this face image acquisition request is input, the host device 2 starts the face image acquisition process shown in FIG. First, the time stamp added to the moving image of the person who is requested to acquire the face image this time is read (s101). The host apparatus 1 transmits a face image transmission request of the date and time indicated by the read type stamp in s101 from the communication unit 23 to the monitoring camera 1 (s102).

  When the monitoring camera 1 receives the face image transmission request from the host device 2 (s111), it reads the date and time indicated by the time stamp included in the transmission request (s112). The surveillance camera 1 reads from the still image storage unit 17 a face image to which a time stamp closest to the date and time read in s112 is added (s113), and transmits this face image from the communication unit 16 to the host device 2 as a still image. (S114).

  When receiving the face image transmitted from the monitoring camera 1 (s103), the host device 2 displays this face image on the display unit 26 (s104). Then, the host device 2 performs a face matching process for matching the face image transmitted from the monitoring camera 1 this time with the face image of the registrant stored in the registrant data storage unit 22 (s105). This face matching process is performed using a well-known Gabor Wavelet and Graph Matching Method, as in the above embodiment. The host device 2 outputs the result of the face collation process relating to s105 (s106). In s106, the result of face collation is displayed on the display unit 26, or an alarm is given if necessary.

  As described above, since the host device 2 can acquire a face image that can be appropriately face-matched from the monitoring camera 1, security is not lowered. In addition, since the face image used for face matching is transmitted as a still image from the monitoring camera 1 to the host device 2, traffic on the network 3 connecting the monitoring camera 1 and the host device 2 is not increased unnecessarily. Therefore, even in a monitoring system in which many monitoring cameras 1 are connected to the host device 2, the monitoring area can be properly monitored in the host device 2, and security is not lowered.

It is a figure which shows the structure of the monitoring system which is embodiment of this invention. It is a figure which shows the structure of the surveillance camera applied to the surveillance system which is embodiment of this invention. It is a figure which shows the structure of the host apparatus applied to the monitoring system which is embodiment of this invention. It is a flowchart which shows the moving image process in the surveillance camera of this embodiment. It is a flowchart which shows the face image transmission process in the surveillance camera of this embodiment. It is a flowchart which shows the face image reception process in the host device of this embodiment. It is a flowchart which shows the moving image acquisition process in the host device of this embodiment. It is a flowchart which shows the moving image transmission process in the surveillance camera of this embodiment. It is a figure which shows the structure of the surveillance camera applied to the surveillance system which is another embodiment of this invention. It is a figure which shows the structure of the host apparatus applied to the monitoring system which is another embodiment of this invention. It is a flowchart which shows the moving image transmission process in the surveillance camera of another embodiment. It is a flowchart which shows the face image storage process in the surveillance camera of another embodiment. It is a flowchart which shows the moving image reception process in the host apparatus of another embodiment. It is a flowchart which shows the face image acquisition process in the host apparatus of another embodiment. It is a flowchart which shows the face image transmission process in the surveillance camera of another embodiment.

Explanation of symbols

1-surveillance camera 2-host device 3-network 12-imaging unit 13-FF module 14-image compression unit 15-moving image storage unit 16-communication unit 17-still image storage unit 22-registrant storage unit 23-communication unit 24 -Face collation unit 25-Image expansion unit 26-Display unit 27-Face image storage unit 28-Movie storage unit

Claims (9)

In a surveillance system in which multiple surveillance cameras are connected to a host device via a network ,
The surveillance camera is
Imaging means for imaging the surveillance area;
A moving image storage unit that stores a captured image of the monitoring area captured by the imaging unit as a moving image with a time stamp indicating an imaging date and time ;
If a person is imaged in the captured image of the monitoring area imaged by the imaging means, face image extraction means for extracting the person's face image;
Time stamp adding means for adding a time stamp indicating the imaging date and time of the captured image obtained by extracting the face image to the face image extracted by the face image extracting means;
The face image cut-out unit cutout, and a still picture transmission means for transmitting to said host system in a still image a face image in which the time stamp adding unit has added a time stamp,
The host device is
For a face image transmitted as a still image from the surveillance camera, when a moving image acquisition request for this face image is input, transmission of the moving image of the date and time indicated by the time stamp added to the face image is performed. Provided with a video request means for requesting a surveillance camera;
Furthermore, the surveillance camera is
Upon receiving the video transmission request from the previous SL upper apparatus, the predetermined time of moving across the time indicated by the time stamp included in the received transmission request, read out from said video storage means, to said upper device It has a video transmission means to transmit ,
Monitoring system. The face image cutout unit of the monitoring camera stores a plurality of frames in a buffer of the captured image in which the person is captured if a person is captured in the captured image of the monitoring area captured by the imaging unit And selecting a captured image from which the face image is cut out from the captured images of a plurality of frames stored in the buffer.
The monitoring system according to claim 1. The moving image storage means of the monitoring camera stores a plurality of frames of captured images stored in the buffer as moving images of the monitoring area.
The monitoring system according to claim 2. The host device is
Face image storage means for storing a registrant's face image;
A face collation unit that collates a face image transmitted as a still image from the monitoring camera with a registrant's face image stored in the face image storage unit and outputs the result. , according to claim 1 monitoring system The host device is
Face image storage means for storing a face image transmitted as a still image from the monitoring camera;
Display means for displaying the face image stored in the face image storage means,
The moving image requesting unit receives an input relating to a moving image acquisition request for the face image displayed by the display unit.
The monitoring system in any one of Claims 1-4 . Imaging means for imaging the surveillance area;
Moving image storage means for storing a captured image of the monitoring area captured by the imaging means as a moving image;
If a person is imaged in the captured image of the monitoring area imaged by the imaging means, face image extraction means for extracting the person's face image;
Time stamp adding means for adding a time stamp indicating the imaging date and time of the captured image obtained by extracting the face image to the face image extracted by the face image extracting means;
The host unit connected via a network, cutout said face image cut-out unit, a still picture transmitting means for the time stamp adding means to send a still image a face image to which a time stamp,
Upon receiving the video transmission request from the previous SL upper apparatus, the predetermined time of moving across the time indicated by the time stamp included in the received transmission request, read out from said video storage means, to said upper device A surveillance camera comprising: a moving image transmitting means for transmitting. In a surveillance system in which multiple surveillance cameras are connected to a host device via a network ,
The surveillance camera is
Imaging means for imaging the surveillance area;
A moving picture transmitting means for sending a picked-up image of the monitoring area picked up by the image pick-up means to the high-order apparatus as a moving picture to which a time stamp indicating the image pickup date and time is added ;
If a person is imaged in the captured image of the monitoring area imaged by the imaging means, face image extraction means for extracting the person's face image;
A time stamp adding means for adding a time stamp indicating the imaging date and time of the captured image obtained by extracting the face image to the face image extracted by the face image extracting means;
Still image storage means for storing the face image with the face image cutout means cut out and the time stamp adding means with the time stamp added as a still image;
The moving image transmitting means transmits the moving image as an image coarser than the face image stored in the still image storage means,
In addition, the host device is
When a face image acquisition request is input for a person imaged in the moving image transmitted from the surveillance camera, the date and time indicated by the time stamp added to the position where the person is imaged in the moving image is input . Still image request means for requesting the surveillance camera to transmit a face image,
Furthermore, the surveillance camera is
Upon receiving the transmission request of the face image from the previous SL upper apparatus, a face image closest timestamp date indicated by the time stamp included in the received transmission request is added, from the still image memory means Comprising a still image transmitting means for reading and transmitting to the host device ;
Monitoring system. The face image cutout unit of the monitoring camera stores a plurality of frames in a buffer of the captured image in which the person is captured if a person is captured in the captured image of the monitoring area captured by the imaging unit And selecting a captured image from which the face image is cut out from the captured images of a plurality of frames stored in the buffer.
The monitoring system according to claim 7. Imaging means for imaging the surveillance area;
A moving image transmitting unit that transmits a captured image of the monitoring area captured by the imaging unit to a higher-level device connected via a network in a moving image with a time stamp indicating the imaging date and time ;
If a person is imaged in the captured image of the monitoring area imaged by the imaging means, face image extraction means for extracting the person's face image;
Time stamp adding means for adding a time stamp indicating the imaging date and time of the captured image obtained by extracting the face image to the face image extracted by the face image extracting means;
Still image storage means for storing the face image with the face image cutout means cut out and the time stamp adding means with the time stamp added as a still image;
Upon receiving the transmission request of the face image from the previous SL upper apparatus, a face image closest timestamp date indicated by the time stamp included in the received transmission request is added, from the still image memory means A still image transmission means for reading and transmitting to the higher-level device,
The video transmitting means transmits a coarse image than the face image storing moving the still image memory means, a surveillance camera

Images