JP2006287893A - Network camera system - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide a network camera system capable of ensuring security without notifying the other person of the situation in the office or house. <P>SOLUTION: The present invention relates to the network camera system wherein video information of a monitor camera is transmitted from external network connection equipment such as a router to a monitor center via a network communication network such as WAN or the Internet so as to execute monitoring and remote control in real time. The monitor camera can perform real-time processing in an analysis means P which compares a previous frame of video with a current frame to extract only a changed portion, a composition means Q which pastes the video analyzed and extracted by the analysis means on another image or video, and a compression means R which compresses the video composed by the composition means and sends it to the monitor center 5 via the network communication network such as WAN or the Internet. <P>COPYRIGHT: (C)2007,JPO&INPIT

Description

  The present invention relates to a network camera system for ensuring security against unauthorized intruders in, for example, an office or a house by a network communication network.

Conventionally, as disclosed in, for example, Patent Document 1, as a network camera system, intrusion information sensed by an installed sensor is input to a surveillance camera without being noticed by an unauthorized intruder, and the image and message are input to a WAN. And a remote security monitoring system using a monitoring camera and a mobile phone having a function of automatically sending mail to a plurality of mobile phone mail addresses designated via a network communication network such as the Internet.
JP 2004-31650 A

  However, according to the conventional surveillance camera, the background part is simultaneously displayed simultaneously with the changed part of the image / video by the intruder, so that the situation in the office or in the house can be easily known to others. Could not be secured sufficiently.

  Therefore, the present invention was created in view of the existing circumstances such as the above, and by extracting only the changed part of the image and video so as not to give the background, the situation in the office or house can be It is an object of the present invention to provide a network camera system that can ensure security without notifying others.

  In order to achieve the above object, in the present invention, monitoring and remote control are executed in real time by transmitting video information of a monitoring camera to a monitoring center via a network communication network such as a WAN or the Internet network. In the network camera system, the surveillance camera compares the previous frame of the video with the current frame, extracts only the changed portion, and pastes the video analyzed and extracted by the analysis to another image or video. The combining unit and the compression unit that compresses the video synthesized by the synthesizing unit and sends the compressed image to the monitoring center via the network communication network are made real-time processable.

  Further, the synthesizing means can make an image in which only the changed video analyzed and extracted by the analyzing means is pasted and synthesized with the replaced still image.

  Further, the synthesizing means can make an image in which an image unchanged by the analyzing means is replaced with a pseudo image.

  The synthesizing means may include face recognition processing means for projecting the peripheral portion of the face recognized by the video displayed on the surveillance camera on the video.

  In addition, the monitoring camera can operate the monitoring camera by detection of a human detection sensor such as an infrared sensor or a far-infrared sensor to enable photographing, recording, and network distribution.

  Connect the camera and video device to a personal computer via USB, video capture hardware, etc., blindly synthesize by motion detection by analysis means, face recognition by face recognition processing means, and distribute to network communication networks such as WAN and Internet It can be provided with software that is processed by a personal computer.

  In the network camera system according to the present invention configured as described above, the analysis means compares the previous frame and the current frame of the video, that is, the temporally continuous image, and extracts only the changed portion, Let the video detect motion. In addition, the compositing means only combines the changed video that has been analyzed and extracted by the analyzing means with the replaced still image, so that the video that is not originally intended to be disclosed to outsiders is deleted. It enables remote monitoring via a network while protecting privacy. Further, the synthesizing unit makes it possible to change the image that has not been changed by the analyzing unit into a video that is replaced with a pseudo image, and to allow the pseudo image to be pasted with another image or video.

  According to the present invention, by extracting only the changed part of the image / video so as not to give a background, security can be ensured without notifying other persons of the situation in the office or in the house.

  That is, the present invention is a network camera system in which monitoring and remote control are performed in real time by transmitting video information of a monitoring camera to a monitoring center via a network communication network such as a WAN or the Internet network. The camera includes an analysis unit that compares the previous frame of the video with the current frame and extracts only the changed portion, a synthesis unit that combines the video analyzed and extracted by the analysis unit with another image or video, and the synthesis unit This is because the compression means that compresses the synthesized video and sends it to the monitoring center via the network communication network can be processed in real time, so that only the motion-detected video is pasted on the replaced still image. By combining and synthesizing the video, it is possible to delete the video that you do not want to be disclosed to outsiders. While protecting the Bashi it is possible to network monitoring.

  In addition, since the synthesizing means combines only the changed video analyzed and extracted by the analyzing means with the replaced still image, the synthesized image is easily combined with the still image. Can be concealed.

  Further, since the synthesizing unit uses the image that has not been changed by the analyzing unit as a video replaced with the pseudo image, the video that is not originally intended to be disclosed to an outsider can be easily hidden by the pseudo image.

  The synthesizing means includes face recognition processing means for displaying the peripheral portion of the face recognized by the video displayed on the monitoring camera on the video, so that when the face is recognized on the monitoring camera video, the periphery of the face in real time The part can be projected on the video and the intruder can be caught.

  In addition, the surveillance camera operates the surveillance camera by the detection of a human detection sensor such as an infrared sensor or far-infrared sensor, so that photographing, recording, and network distribution are possible. For example, when a person entering or exiting an automatic door is detected. In addition, it is possible to easily construct an automatic monitoring network system in which security can be ensured by a monitoring camera that is automatically set up on the network.

  Furthermore, distribution by a network camera is currently made to an unspecified number on the Internet or the like. In this case, the camera zoom at a high magnification causes the passerby's face to be disclosed to the public and infringes on privacy. At this time, privacy can be ensured by mosaicing and blinding the face of the camera image by the face recognition processing means. In particular, in the case of a videophone, except for face detection by the face recognition processing means and motion detection by the analysis means, the mosaic is effective for indoor private confirmation.

  Also, software that connects the camera and video device to a personal computer via USB, video capture hardware, etc., blindly synthesizes by motion detection by analysis means, face recognition by face recognition processing means, and processes network distribution on a personal computer. Therefore, blind synthesis output using a personal computer is possible by incorporating a blind function in a general surveillance camera other than the purpose of network distribution into the surveillance camera or by attaching it externally.

  In the case of a local area network, for example, when a blind camera for face recognition is attached to a nursing / medical bed, the privacy of the patient can be ensured, and the nurse work is not affected by the ON / OFF switch of the blind at the nurse center.

  Furthermore, when monitoring with a regular monitor in a public facility, the monitor output can be blinded in the face, and in an emergency, the blind can be turned on and off at the monitoring center. Coexistence is possible.

  Hereinafter, the best mode for carrying out the present invention will be described with reference to the drawings. Reference numeral 1 shown in the figure is a monitoring camera in the network camera system according to the present invention. Monitoring and remote control are executed in real time by transmitting from the external network connection device 2 to a monitoring center 5, a PC at work, a cellular phone (PDA), etc. via a network communication network 3 such as a WAN or the Internet. I am trying.

  In other words, this system inputs images in an office or house to a plurality of monitoring cameras 1 such as a CCD camera or an infrared camera capable of capturing 30 frames per second managed by an internal address, and monitors them. Monitoring is performed by transmitting the video information of the camera 1 from an external network connection device 2 such as a router to a workplace PC, a mobile phone PDA, and further to a monitoring center 5 via a network communication network 3 such as a WAN or the Internet. Remote control is made.

  As shown in FIG. 1, the monitoring camera 1 includes an analysis means P that enables motion detection by comparing the previous frame and the current frame of the video of the monitoring camera 1 and extracting only the changed portion, and analysis. Combining means Q for pasting the video analyzed and extracted by means P to another image or video, and compressing the synthesized video by arbitrary encoding, and then via a network communication network 3 such as WAN or Internet network Compression means R for sending out, and each means P, Q, R can be processed in real time.

  This synthesizing means Q is a video obtained by pasting and synthesizing only the changed video analyzed and extracted by the analyzing means P with the replaced still image, or an image not changed by the analyzing means P as a pseudo image. It is also possible to replace the video. Further, an image having no motion is stored in the CPU 4 as basic data, and the basic data is periodically updated by determining the brightness of the image at this time.

  As a specific configuration of the monitoring camera 1, as shown in FIG. 2, the monitoring camera 1 has high functionality in which a network camera is equipped with an STB (set top box) function which is a bidirectional multimedia communication terminal. It is supposed to be. In addition, when using the existing surveillance camera 1 and video camera, a video signal and an audio signal may be connected to this STB.

  The input / output terminals of the surveillance camera 1 include NTSC composites, voice microphones, Ethernet (registered trademark) LAN terminals, privacy SW, external network SW, power SW, on / off switch of electric light / TV, intercom for call conversation Etc. are installed. Furthermore, the sensor may be either wireless or wired.

  The intercom for call conversation is operated by monitoring and remote control of a work PC or mobile phone PDA. The composite and voice microphones are connected to a TV phone, VTR for recording original images, etc., and the Ethernet (registered trademark) LAN terminal, the light / TV on / off switch, and the intercom for calling conversation are monitored via the home network. Each floor is operated by STB monitoring, camera settings, sensor control, etc., and multiple recording can be performed on a LAN disk. At this time, the router 2 itself is connected to devices such as a workplace PC and a mobile phone PDA via a network communication network 3 such as a WAN or the Internet network.

  Next, the network processor function of the STB will be described. The STB has a function of processing a chroma key processing of a video signal without using a blue back and a function of always distributing a pseudo video on a network. This pseudo image can be switched remotely, but cannot be switched remotely when the STB manual switch is OFF. At this time, the video itself is managed with an ID password and is constantly distributed to the monitoring center 5 and others. Moreover, the pseudo image can be both a moving image and a still image, and can be inserted from an external device.

  In addition, the STB uses the network terminal of the surveillance camera 1 to perform control settings from inside and outside the home. Furthermore, the STB is connected to a TV, a monitor TV or the like so that all operation settings can be performed by remote control using a remote controller. In addition, it is possible to control a light, a TV, a curtain, etc. by selecting a chroma key process at the same time as a TV telephone function, incorporating a recording device, and judging the brightness of a camera image. In addition, the STB can perform so-called wireless recording in which the recording medium is hidden in an arbitrary place on the LAN.

  At the time of occurrence of an abnormality in which it is determined by the analyzing means P that there is a motion in the video, this abnormality is notified to the monitoring center 5 simultaneously with the alarm notification 6, and the monitoring center 5 determines the situation and reports it to a predetermined place. In addition, the surveillance camera 1 attaches a wireless device tag so that the movement by the analysis means P is not detected and is ignored, and only a moving intruder is detected and displayed. .

  The synthesizing unit Q includes a face recognition processing unit S for projecting the peripheral portion of the intruder's face displayed on the monitoring camera 1 and recognized by the video extracted and analyzed by the analyzing unit P on the video. For example, the face recognizing processing means S, when combining the video with another image or video in the synthesizing means Q, makes sure that the recognized intruder's face is clearly projected before the monitoring center 5, workplace, etc. To a PC, a mobile phone (PDA), etc. In addition, for the purpose of privacy protection, the face of the intruder of the video may be concealed with, for example, a mosaic, a mask, or a crushed video. This may be a general surveillance camera.

  In the combining means Q, as shown in FIG. 3, the monitoring image before combining and the normal monitoring screen with blind, normal and filtered can be switched remotely with each other. In addition, the monitoring screen that records the occurrence of an intrusion abnormality and the monitoring screen at the time of anomaly that combines motion detection images on the normal monitoring screen that is blinded, normal, and filtered can be switched remotely. Yes. The blind, normal, and filter may be moving images, still images (white and black), raster images, vector images, and the like. In addition to the fitting of the frame, the blind includes mosaic and stained glass tone in real time by image processing.

  The face recognition processing means S described above may output only the face of the intruder with this face recognition function and make the others blind, or conversely blind only the face and output others. Further, as a specific example of the blind, for example, a clock is transmitted as a part of the video, thereby making it easy to specify the home or the target location. At this time, the transmission part can be moved in the vertical and horizontal directions in real time. In addition, the blind type can be easily changed by remote control from the monitoring center 5, a PC at work, a mobile phone (PDA), or the like.

  In the case of a videophone, except for face detection by the face recognition processing means S and motion detection by the analysis means P, the mosaic is effective for private confirmation in the room. For example, a peripheral portion when a face is recognized in the monitoring camera image is displayed on the image to catch an intruder. Further, in order to confirm the video location, the image quality, and the live output, a part of the video is made transparent and translucent (the clock part in FIG. 8). That is, it is possible to specify the home and target location by transmitting a clock that is a part of the video. This transparent part can be moved up and down and left and right in real time, and the blind type can be changed by remote control. In addition, as the types of blinds, as shown in FIG. 8, for example, mosaic, out-of-focus focus, stained glass or the like is used for the video. Examples of normal video, face mosaic video, and mosaic video other than the face are shown in FIG. 9 as privacy maintenance in the case of a videophone.

  As privacy protection by face recognition, a blind image when a face is recognized by a live camera is shown in FIG. In other words, in the past, monitoring at public facilities has no privacy, so by recognizing and blinding the face in this way, live camera images at public facilities can be safely released, zooming up without considering privacy, Recognized faces can be rewritten into animations, portraits, solid images, and so on.

  Further, as a measure against home position deviation of a non-fixed camera, a marker is arranged and set, a home position by marker storage is detected, and the camera is panned, tilted and zoomed to a predetermined position. Markers include wall clocks, stuffed animals, posters, and color panels. In particular, face recognition using posters is highly effective.

  The image displayed on the surveillance camera 1 is a moving image shot with no humans or animals coming and going, but it is a repetition of the same image. This is a surveillance video that you do not want to show to outsiders. In this case, an image in which no motion is detected is replaced with a pseudo image and output. For example, as shown in FIG. 4, an arbitrary still image described as “in security” is pasted on a motionless image projected by the monitoring camera 1. At this time, it can be confirmed that the power supply SW of the monitoring camera 1 is ON and is operating normally.

  In addition, the synthesizing means Q can be replaced with a stored moving image in addition to the normal monitoring screen on which the filter is applied. For example, it can be replaced and combined with a landscape video. By only combining the motion-detected video with the replaced still image and combining it, the video is monitored from the network while protecting the privacy in a state where the video that is not originally desired to be disclosed to the outside is deleted. It becomes possible.

  Such replacement video data can be output from an external video input, a STB memory in a surveillance camera, a memory that can be replaced and inserted such as a compact flash (registered trademark), and the like. Further, the monitoring camera 1 can be connected to a VTR, and video images other than those for monitoring purposes can be output. Furthermore, the replaced video can also be used to picture-in-picture the abnormal part of the intrusion video. For example, FIG. 5 is a composite image in which only an abnormal part is captured in the replaced moving image. FIG. 6 is an enlarged view of the motion detection area, and when the motion detection is accurately captured at this time, the background is also expanded together with the back background as a measure for preventing a screen loss. In some cases, the upper body moves and the lower body does not move. In this case, the human body region is enlarged by using the face recognition function together to recognize the entire movement.

  The surveillance camera 1 has a function of periodically resetting and retrieving the reference background by setting the overall brightness change criterion as a measure for preventing the brightness change of the entire image due to sunlight. In that case, the analysis means P also confirms a partial movement change. Further, in the case of the dusk due to the judgment of the input luminance of the monitoring camera 1, it has a function of outputting the output signal of the monitoring camera 1 and turning on the light switch, TV switch, etc. with the remote control. In this way, it is effective to prevent nighttime intrusion by providing a function for automatically turning on an electric lamp, TV, and the like.

  In addition to motion detection by the analysis means P, the surveillance camera 1 itself is linked with a human detection (human sensor) sensor such as an infrared sensor or a far infrared sensor, and when an intruder is detected by the human detection sensor. The surveillance camera may be automatically operated to perform shooting, recording, network distribution, and the like. At this time, sensors, alarms, serial, make contacts, and external input / output are also possible. In this system, alarms are issued according to the level of crisis management. For example, it is level 1 for a registered face recognition person, level 2 for motion detection, level 3 for human detection sensor, level 4 for face detection, level 5 for synthesis of motion detection and human detection sensor, and the like. In addition, it has a function to prevent malfunction by comprehensive judgment processing of face recognition, far-infrared sensor, and motion detection. At this time, the registered person is not detected when the face recognition function is used. Moreover, far-infrared sensors do not detect pets. For example, possible malfunctions in the case of motion detection include, for example, subtle motions such as changes in illuminance, aquarium waves, and wind chimes. Further, as a possible malfunction of the human sensor, wearing a wet suit basically does not activate the far infrared sensor, but the far infrared sensor has a false detection of a pet. Further, as a possible malfunction of face recognition, in the case of wearing a mask, it is basically not recognized.

  At this time, software that allows external video equipment such as a USB camera and a composite output camera to be connected to a personal computer, performs blind synthesis by motion detection by the analysis means P, face recognition by the face recognition processing means S, and processes network distribution on the personal computer. I have.

  In addition, the blind function of a general monitoring camera 1 other than the purpose of network distribution is built into the monitoring camera 1 or is externally attached, thereby enabling blind composite output using a personal computer.

  In addition, in the case of a local area network, for example, when a blind camera for facial recognition is attached to a nursing / medical bed, the privacy of the patient can be ensured, and the nurse work is not affected by the ON / OFF switch of the blind at the nurse center. I have to.

  Moreover, in the monitoring by a normal monitor in a public facility, the monitor output can be blinded in the face, and in an emergency, the blind can be turned on / off at the monitoring center. It is designed to be compatible.

  Furthermore, the monitoring camera 1 may be normally only for photographing and recording, and when the intruder is detected by the human detection sensor, network distribution may be automatically performed by the monitoring camera.

  In addition, the surveillance camera 1 is normally turned off or in a sleep state, and when the intruder is detected by the human detection sensor, the surveillance camera 1 is automatically turned on to shoot, record, and distribute the network. May be performed. For example, when an automatic door open / close detection sensor or the like works, the surveillance camera 1 may be automatically turned on in conjunction with it to enable monitoring of the entering / exiting.

  In addition, when using the face recognition function by registering a person in advance, it can be set so that the registered person is not detected by the human detection sensor.

  Next, an example of the operation of the best mode configured as described above will be described based on the flow transition shown in FIG. Motion is detected by the analyzing means P from the video of the monitoring camera 1 and various sensors, and an image without motion is stored in the CPU 4 as basic data, which is replaced with a pseudo image. At this time, the basic data is periodically updated by determining the contrast of the image.

  The PC or mobile phone is set to be always connected to the network, connected to the network as necessary, or connected to the network via the monitoring center 5. In the case of original image recording, the location of the VTR cannot be specified by a third party, which is highly effective. This is also effective for remote recording over a network communication line and prevention of arson.

  On the other hand, when a moving part is generated in the video by an intruder or the like, the moving part is extracted by the analyzing unit P, and only this extracted part is pasted on the pseudo video by the synthesizing unit Q. In parallel with this, the alarm notification 6 and the monitoring center 5 are informed of this abnormality at the same time, and the monitoring center 5 judges the situation and reports it to a predetermined place. At this time, billing is performed at the same time.

  Further, when the image is combined with another image or image by the compositing means Q by the face recognition processing means S, the face of the recognized intruder is clearly projected, and then the monitoring center 5 or a PC such as a workplace is used. To a mobile phone (PDA). Alternatively, only the face of the intruder may be output by this face recognition function and the others may be blinded, or only the face may be blinded and the other may be output as an image.

  Then, the video on the network and the original image of the video on the network are simultaneously output and transmitted to a PC or mobile phone at a remote location, and confirmation and control are performed by the PC or mobile phone. At this time, the monitoring center 5 also notifies the PC or mobile phone. After confirmation by the PC or mobile phone, control is performed from the PC or mobile phone so as to synthesize the pseudo image and the moving part by the synthesizing means Q, or to interrupt the pseudo image and switch to the original image.

  Next, the operation of the network camera system using a personal computer will be described with reference to the flowchart of FIG. First, in step 1, the HTTP server is activated. In step 2, the blind screen is recorded in an area that can be opened to the public so that it can be accessed from the network. Capture the image in step 3. In step 4, the moving body detection process is performed by the analyzing means P. In step 5, it is determined whether or not there is moving object detection. If there is moving object detection, the process proceeds to step 6, and a difference range is cut out in a rectangular shape from the captured image. Next, the blind screen is redrawn in step 7, and the rectangle cut out in step 8 is displayed on the blind screen. That is, the image stored in the area that can be disclosed to the outside is updated. In step 9, the changed image is saved on the disk. The image having changed in step 10 is uploaded to the FTP server (depending on the setting). At step 11, the image that has changed is sent as an attached mail (depending on the setting). In step 12, alarm notification processing to the security company is performed. When the mobile phone or the PC is connected in step 13, alarms and live distribution are executed and recorded, then the process returns to step 3 and the subsequent operations are repeated.

  On the other hand, if it is determined in step 5 that there is no moving object detection, face recognition processing is performed by the face recognition processing means S in step 14, and it is determined in step 15 whether or not there is face recognition. In step 16, the face range is cut out in a rectangular shape from the captured image, and the process proceeds to step 7 to redraw the blind screen, and the process proceeds to step 8 and the subsequent operations are repeated. If there is no face recognition in step 15, the process returns to step 3 and the subsequent operations are repeated.

It is explanatory drawing which shows the system flow transition for operation | movement description in the best form for implementing this invention. It is a block diagram which similarly shows the function of a system. It is explanatory drawing which shows an example of a synthetic | combination means. It is explanatory drawing which similarly shows the other Example of a synthetic | combination means. It is explanatory drawing which shows the further another Example of a synthetic | combination means similarly. It is explanatory drawing which shows the further another Example of a synthetic | combination means similarly. It is a flowchart figure similarly explaining the synthetic | combination procedure by a personal computer. It is explanatory drawing which similarly shows the kind of blind. It is explanatory drawing which shows the specific example of privacy maintenance in the case of a videophone similarly. It is explanatory drawing which shows the specific example of a blind image when a face is similarly recognized with the live camera.

Explanation of symbols

P Analyzing means Q Combining means R Compression means S Face recognition processing means 1 Surveillance camera 2 External network connection device such as router 3 Network communication network such as WAN or Internet network 4 CPU
5 Monitoring Center 6 Alarm Alarm

Claims (6)

  A network camera system in which monitoring and remote control are performed in real time by transmitting video information of a monitoring camera to a monitoring center via a network communication network such as a WAN or the Internet network. Means for comparing the current frame with the current frame and extracting only the changed portion, combining means for pasting the video analyzed and extracted by the analyzing means to another image or video, and compressing the video synthesized by the combining means A network camera system characterized in that the compression means for sending to the monitoring center via the network communication network can be processed in real time.   The network camera system according to claim 1, wherein the synthesizing unit is an image obtained by pasting and synthesizing only the changed video analyzed and extracted by the analyzing unit with the replaced still image.   The network camera system according to claim 1, wherein the synthesizing unit uses a video obtained by replacing the image without change by the analyzing unit with a pseudo image.   The network camera system according to claim 1, wherein the synthesizing unit includes a face recognition processing unit that displays a peripheral portion of the face recognized by the video displayed on the surveillance camera on the video.   The network camera system according to any one of claims 1 to 4, wherein the surveillance camera is capable of photographing, recording, and network distribution by operating the surveillance camera by detection of a human detection sensor such as an infrared sensor or a far infrared sensor. Software that allows camera and video equipment to be connected to a personal computer via USB, video capture hardware, etc., blindly synthesizes by motion detection by analysis means, face recognition by face recognition processing means, and allows network distribution to be processed by a personal computer The network camera system according to claim 5.