JP2009010579A - Video intercom device - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide a video intercom device capable of performing stable recognition and tracking control of a visitor by optimizing imaged video for image analysis processing and imaged video to be transmitted to a master unit. <P>SOLUTION: The video recognizing part 102 of a slave unit 1 provided to the video intercom device receives video photographed and output by a camera part 11 at a period following a first period with pixels thinned and analyzes the presence/no presence of a visitor on the basis of the received video. A visitor determining part 103 determines the presence/no presence of the visitor on the basis of the result of recognizing the video. The video photographed and output by the camera part 11 is transmitted to the master unit at a period following a second period shorter than the first period. Since this image processing not only improves detection accuracy by the video recognizing part 102 using a complicated algorithm requiring a processing time but also performs temporal filter processing of a moving object, staggering of positional information of the visitor caused by subtle shaking of the visitor can be prevented. <P>COPYRIGHT: (C)2009,JPO&INPIT

Description

  The present invention relates to a television interphone device, and more particularly to a television interphone device that can automatically recognize a visitor and perform electronic zoom and pan / tilt operations.

  For example, as disclosed in Patent Document 1, a television intercom apparatus that performs a pan / tilt operation has been conventionally known. This TV intercom device is equipped with an interphone entrance unit with camera and an interphone room base unit with a monitor that is connected to the interphone entrance unit with camera in a wired manner. On the other hand, the pan / tilt operation is realized by cutting out a part of an arbitrary imaging range of the intercom entrance with a camera and performing an enlargement process.

  Further, as disclosed in Patent Document 2, tracking control is performed when a part of the imaging range of the interphone entrance with a camera is cut out, and a visitor's face or the like is automatically monitored. There is known a TV intercom device that displays in the center of a monitor of a room master.

  Further, as disclosed in Patent Document 3, a moving body is automatically detected by dividing an imaging region into blocks and performing a difference process with a past captured image in units of blocks, and an interphone entrance with a camera There is known a television intercom apparatus that controls the cut-out of a partial area of an imaging range by a machine and displays a moving body in the center of the monitor of the monitor intercom base unit.

  According to the TV intercom apparatus having such a configuration, a request for locally expanding and confirming a visitor's face in the house and a request for confirming the overall appearance of the front door including the visitor, It is possible to meet together without performing troublesome operations.

  However, in the TV intercom apparatus having such a configuration, the position information of the visitor centered on the face is recognized in real time by the video signal of the video taken by the camera intercom entrance cordless handset, and the position information is monitored. Although it is necessary to transmit to the living room master unit, in the case of an interphone entrance unit with a camera with limited processing capability, a device for stably performing visitor recognition and tracking control is required.

  In order to stably perform visitor recognition, in the configuration of Patent Document 1, it is necessary for the customer to manually perform the tracking operation while watching the monitor video of the monitor-equipped intercom room master unit. In the configuration of Patent Document 2, tracking control is automatically performed by image analysis of captured images. However, a home visitor manually searches for a visitor while watching a monitor image of a monitor intercom room master unit, and specifies a tracking target in advance. It is necessary to. In the configuration of Patent Document 3, real-time tracking control is achieved by image analysis of captured images. However, since all moving objects are subject to tracking control regardless of human beings, a visitor must always be a monitor intercom parent with monitor. It may not be possible to display on the monitor of the machine.

In addition, as a means for detecting a human regardless of image analysis of a captured image, a heat source movement detection element typified by a pyroelectric sensor that detects a change in far-infrared rays is conventionally used. Restrictions such as being unable to correctly recognize a person unless the condition that the person being detected is constantly moving or having a temperature difference from the background is satisfied, and position information within the detection area of the sensor cannot be obtained Due to the condition, there is a problem of malfunction in automatic detection of visitors before the call button is pressed.
JP 2004-304558 A JP-A-6-38214 JP-A-11-275568

  Therefore, there is a problem to be solved in terms of optimizing the captured image for image analysis processing in the interphone entrance device with camera and the captured image to be transmitted to the interphone room master unit with monitor. The brightness of the subject obtained from the sensitivity value set by the AGC (automatic sensitivity adjustment) function of the photographing unit of the intercom entrance unit with camera by performing image analysis control on the intercom entrance unit with camera according to the installation environment The information is also determined based on the information on the presence / absence of the heat source obtained by the heat source movement detection element, the temperature information obtained by the temperature measurement, or the date information obtained by the time measuring function in combination with the image analysis result of the captured image. There is a problem to be solved.

  The present invention has been made to solve these problems, and the object of the present invention is to optimize the captured video for image analysis processing and the captured video to be transmitted to the master unit, thereby stabilizing the visit. It is to provide a television intercom device capable of recognition and tracking control of a person.

  A TV intercom apparatus according to the present invention is a TV interphone apparatus including a slave unit having a camera unit provided at a position where a visitor can be photographed, and a plurality of master units having a display unit communicating with the slave unit. The slave unit further inputs a video imaged and output by the camera unit in a cycle according to the first cycle, and performs video recognition means for analyzing presence / absence of a visitor based on the input video, and a video recognition result A visitor judging means for judging the presence or absence of a visitor based on the above, and a video transmitting means for inputting a picture taken and output by the camera unit at a period according to the second period and transmitting it to the master unit Have. The master unit further includes a received video display unit that receives the video transmitted by the video transmission unit and displays the received video on the display unit.

  Further, the slave unit inputs the video imaged and output by the camera unit with the first pixel configuration, and converts the input captured video information into the second pixel configuration based on the video information obtained by the visitor. Video recognition means for analyzing presence / absence, visitor determination means for determining presence / absence of a visitor based on the result of video recognition, and input and input video captured and output by the camera unit with the first pixel configuration It is preferable to have video transmission means for converting the captured video information into the third pixel configuration and transmitting it to the master unit.

  Preferably, the slave unit further includes means for accumulating and storing the analysis results obtained by the video recognition means for a certain period.

  Preferably, the slave unit further includes an imaging condition acquisition unit that inputs video imaging conditions in the camera unit to the video recognition unit.

  Preferably, the subunit | mobile_unit has further the time-counting means which can measure a month day time.

  Preferably, the subunit | mobile_unit has further the heat-source movement detection part which can detect the motion of the infrared rays which generate | occur | produce from a visitor.

  Preferably, the subunit | mobile_unit has further the temperature measurement means which can measure the air temperature of an installation place.

  According to the present invention, in the video recognition control of a visitor by image analysis, the processing time is reduced by using the frame-removed video obtained by thinning the captured video taken by the intercom entrance with a camera in time. Not only can the detection accuracy be improved by video recognition means using the necessary complex algorithms, but it will also be a temporal filter process for moving objects, so that fluctuations in the visitor's location information due to subtle shaking of the visitor can be avoided. Can be suppressed. On the other hand, even if the transmission path to the master unit is unstable when the transmission path to the master unit is unstable because the transmission path to the master unit is greatly limited by bandwidth such as wireless, the intercom entrance unit with camera is unstable. It is possible to stably acquire the location information of the visitor by the image recognition means.

  In addition, in the video recognition control of a visitor by image analysis, the analysis load is reduced by using the video having a video pixel configuration obtained by thinning out the captured video taken by the intercom entrance with a camera. As a result, not only detection control by inexpensive video recognition means with low analysis capability is possible, but also spatial filtering processing for moving objects, so fluctuation of visitor's location information due to subtle shaking of visitors Can be suppressed. On the other hand, even when the monitor display video pixel configuration of the master unit is constrained, the captured video pixel configuration and the pixel configuration used for the video recognition means that are captured by the intercom entrance with camera are transmitted to the master unit. By using a high-pixel configuration compared to the monitor-display video pixel configuration, when a wide-angle lens is used in the imaging unit of the intercom entrance with a camera, visitors can also use the video recognition means to capture fine movements. It is possible to stably acquire the position information.

  In addition, in the video recognition control of a visitor by image analysis, a unique imaging situation caused by the installation location of the intercom entrance with a camera is generated. In other words, there are areas such as walls that visitors cannot reach within the shooting range, or areas where human traffic other than visitors or cars are always captured at the edge of the screen. It becomes a misrecognition factor by means. The area obtained by excluding such an area from the imaging area of the intercom entrance unit with camera is set as the image recognition target area, and the movement trajectory of the visitor to the operation operation of the call button of the intercom entrance unit with camera is accumulated. By setting the obtained range as the priority video recognition target area, it is possible to stably acquire the location information of the visitor by the video recognition means.

  In addition, by using the correction information of the subject brightness by the AGC (automatic sensitivity adjustment) function in the photographing part of the intercom entrance unit with a camera for the recognition control in the image recognition means, the sensitivity is increased and the image becomes cloudy or dark in the evening. It can be recognized that the subject is being photographed. Increasing the sensitivity increases noise, but by recognizing that the sensitivity has been increased, by setting the threshold for determining the moving body in the image recognition means to be high, the image recognition means Stable acquisition of visitor location information can be performed.

  In addition, by determining whether there is a visitor by combining the date and time information of the location where the intercom entrance with a camera is installed and the information recognized by the video recognition means, it is possible to respond to the season based on the date and time information. It is possible to estimate the visitor's clothes and skin exposure, and correct the recognition information in the video recognition means such as the visitor's face position, etc., so this correction can stably acquire the visitor's position information It can be carried out.

  In addition, the image recognition means makes unstable recognition around the threshold by judging whether or not there is a visitor by combining the heat source movement detection information within the shooting range of the intercom entrance with a camera and the recognition information by the image recognition means. If there is information but the heat source movement detection unit does not detect it, it is judged that the person who is far away or the flickering of vegetation close to the skin color is misrecognized, and conversely, the heat source movement detection unit detects the heat source. However, if the image recognition means recognizes a very high-brightness spot light source, control is ignored as a result of direct irradiation of far-infrared rays, such as sunlight, automobile headlights, and incandescent lamps, onto the heat source movement detector. By performing the operation, it is possible to stably acquire the location information of the visitor.

  In addition, by determining whether or not there is a visitor by combining the temperature information within the shooting range of the intercom entrance unit with camera, the heat source movement detection information, and the recognition information from the image recognition means, the visitor's clothes are determined from the temperature information. The position information of the visitor can be stably acquired by estimating the degree of exposure of the skin and correcting the recognition information by the image recognition means such as the visitor's face position based on this guess. . At the same time, it is possible to perform control for ignoring the erroneous recognition result when the temperature rises to near the human body temperature on a midsummer day, which is a drawback when the heat source movement detection unit is used in combination, based on the temperature information.

Embodiments of the present invention will be described below with reference to the drawings. In the following description, the same parts or components are denoted by the same reference numerals. Since their names and functions are also the same, detailed description thereof will be omitted.
In the TV intercom apparatus of this embodiment, the interphone with a camera that is a device having a function of photographing a visitor and outputting a video signal is called a “child device”, and is a device operated by a resident. The interphone master unit with a monitor that communicates with the slave unit is called “master unit”.

  FIG. 1 shows an overall configuration, FIGS. 2 to 3 show functional configurations, and FIGS. 4 to 7 show flowcharts.

FIG. 1 shows a specific example of the configuration of the television intercom apparatus according to this embodiment.
Referring to FIG. 1, the television intercom apparatus according to the present embodiment is configured to include a child device 1 that is usually installed outdoors and parent devices 2A and 2B that are usually installed indoors. . The subunit | mobile_unit 1 contains the camera part 11, the call button 13, the heat source movement detection parts 14, such as a pyroelectric sensor, the LED illumination 15 which illuminates the imaging region of a camera brightly, the temperature measurement part 16, etc. Base units 2A and 2B include a liquid crystal display unit 21, a call button 22, a recording / playback button 24, an extension call button 25, and the like.

  In the subunit | mobile_unit 1, the camera part 11 is provided in the position which can image | photograph a front door, and is provided in the position which can image | photograph a visitor especially. Similarly, the heat source movement detection unit 14 is provided at a position where the entrance of the entrance can be sensed. In particular, the heat source movement detection unit 14 is provided at a position where the heat source movement within the same range as the imaging range of the camera unit 11 can be detected. The slave unit 1, the master unit 2A, and the master unit 2B communicate with each other through the existing laying line 3, and the operation power of the slave unit 1 is supplied from the master unit 2A through the existing laying line 3.

  2 to 3 are block diagrams showing an internal configuration of each device constituting the television intercom apparatus according to the present embodiment. FIG. 2 shows an internal configuration of the handset 1 of the television intercom apparatus. Base unit 2A and base unit 2B have the same configuration. For this reason, in FIG. 3, the structure of the main | base station 2 which represented the main | base station 2A and the main | base station 2B is shown.

  Referring to FIG. 2, slave unit 1 included in the television intercom apparatus according to the present embodiment calls CPU (Central Processing Unit) 100 that controls the entire slave unit 1, master unit 2 </ b> A, and master unit 2 </ b> B. In the call button 12, the signal transmission / reception unit 107 including a terminal block for attaching the existing line communication line 3 to the base unit 2A, the base unit 2B, etc., the camera unit 11 for photographing a visitor, and the camera unit 11 A recognition video storage unit 101 that stores the captured video signal, a video recognition unit 102 that performs image analysis of whether or not a visitor has been photographed, a visitor determination unit 103 that determines whether there is a visitor, An analysis result storage unit 104 that stores a history of video recognition results in the video recognition unit 102, a date / time storage unit 105 that measures and holds time with a built-in timer function, and a video signal captured by the camera unit 11 It includes a transmission video storage DA (digital-analog) converter 106 that converts a digital video signal to an analog video signal for loading and transmission to the master unit 2 and the like, and a microphone or speaker (not shown). A voice call unit 13 that enables voice calls between the user and the indoor user, a heat source movement detection unit 14 that collects the movement state of the heat source within the imaging range of the camera, and an LED illumination unit 15 that illuminates the imaging range of the camera brightly. The temperature measuring unit 16 that measures the temperature of the place where the handset 1 is installed and the power supply circuit unit 108 are included.

  The CPU 100 is connected to the call button 12, and a call signal is input to the CPU 100 from the call button 12 when the visitor presses it. The CPU 100 includes a camera unit 11, a recognition video storage unit 101, a video recognition unit 102, a visitor determination unit 103, an analysis result storage unit 104, a date / time storage unit 105, and a transmission video storage D /. The A conversion unit 106, the signal transmission / reception unit 107, the heat source movement detection unit 14, the LED illumination unit 15, and the temperature measurement unit 16 are connected, and control signals for controlling them are sent out. When the call signal from the call button 12 is input and the call signal is detected, or the visitor determination unit 103 detects a visitor, the CPU 100 transmits and receives a call start signal, which is a control signal indicating that there is a visitor. The data is sent to the unit 107.

  The signal transmission / reception unit 107 receives the frequency-multiplexed transmission signal transmitted from the parent device 2 via the communication transmission path 3, and converts the transmission signal into the voice signal and the parent device response signal according to the control signal from the CPU 100. Etc., and further to the drive power supply of the slave unit 1.

  The power supply circuit unit 108 is connected to the signal transmission / reception unit 107, creates DC 3.3V, 5V, and 12V necessary for the operation of the slave unit 1 based on the DC 15V received by the signal transmission / reception unit 107, and sets each operation circuit block. Supply power.

  The CPU 100 is connected to the heat source movement detection unit 14, the temperature measurement unit 16, and the month / day / time storage unit 105, and the heat source movement detection unit 14 detects the heat source within the imaging range of the camera as information necessary for the CPU 100 to control. The movement information, the temperature information of the installation location of the child device 1 from the temperature measurement unit 16, and the current date / time information from the date / time storage unit 105 are acquired as needed.

  The camera unit 11 includes a CCD image pickup device having effective pixels of horizontal 1280 × vertical 960 pixels (first pixel configuration), and an image pickup range having a horizontal angle of view of 120 degrees in accordance with a control signal and a time signal from the CPU 100. A video signal picked up by a lens having the above is sequentially transmitted to the recognition video storage unit 101 and the transmission video storage D / A conversion unit 106 in a cycle of 30 screens per second. This video signal is a digital data signal of luminance information and color information for each pixel to which a synchronizing signal for each screen is added.

  The recognizing video storage circuit 101 has a capacity capable of storing the video captured by the camera unit 11 for three screens, and has a ring-shaped memory structure in which new video is overwritten and stored in order from the old video. In accordance with the control signal and the time signal from the CPU 100, the video signal data from the camera unit 11 is reduced into effective pixels of horizontal 320 × vertical 240 pixels (second pixel configuration), and 7 times per second. .Sequentially stored in a cycle of 5 screens (first cycle).

  The video recognition unit 102 reads out the video signal data of two screens delayed by 1 / 7.5 second stored in the recognition video storage unit 101 according to the control signal from the CPU 100, performs image analysis processing, and performs face detection, moving object Three types of detection result flags, detected or not detected, and coordinate data of the detection target are sequentially sent to the visitor determination unit 103.

  The visitor determination unit 103 analyzes the image recognition unit 102 according to the control signal from the CPU 100, the heat source movement information detected by the heat source movement detection unit 14, and the temperature information of the installation location of the child device 1 measured by the temperature measurement unit 16. Based on the above, a process for determining whether there is a visitor is performed.

  The analysis result storage unit 104 has an information storage function that corresponds to the horizontal 320 × vertical 240 coordinates, which is the same as the effective size of the recognition video storage unit 101, and has a visitor when the visitor is recognized. The moving trajectory of the face coordinates is subjected to a weighted calculation of the moving average for each coordinate area, and the calculation result is accumulated and recorded. Since this accumulated record becomes a visitor's face area where the call button 12 can be pressed, the detection process of the face image in the image recognition unit 102 is limited to only within this area, so that the visitor and the passerby are limited. Is being discriminated.

  The heat source movement detection unit 14 is composed of a far infrared pyroelectric sensor that detects movement of a heat source having a temperature difference of 4 ° C. or more from the background, and has a detection distance of 10 m or less. The imaging range of the far-infrared pyroelectric sensor is attached at a position where the horizontal field angle is 120 degrees, which is the same as the imaging range of the camera 11.

  The transmission video storage D / A conversion unit 106 has a capacity that can store the video captured by the camera unit 11 for two screens, and has a ring-shaped memory structure in which the new video is overwritten and stored in order from the old video. The effective pixels centering on the visitor coordinates obtained by the visitor determination unit 103 for the video signal data from the camera unit 11 according to the control signal and the time signal from the CPU 100 are horizontal 640 × vertical 480 pixels ( A rectangular region (third pixel configuration) is cut out, and sequentially stored at a cycle of 30 screens per second (second cycle).

  The transmission video accumulation D / A conversion unit 106 converts the stored digital video signal data into analog data according to the control signal and time signal from the CPU 100, and performs 30 screens per second for the signal transmission / reception unit 107. Sequentially sent in a minute cycle.

  The voice call unit 13 is connected to the signal transmission / reception unit 107, and outputs a voice signal input from the signal transmission / reception unit 107 to an outdoor visitor from a speaker according to a control signal from the CPU 100. The voice uttered by the visitor is acquired, and the acquired voice signal is transmitted to the signal transmitting / receiving unit 107.

  In accordance with a control signal from the CPU 100, the signal transmission / reception unit 107 is a voice signal input from the voice call unit 13, an analog video signal read from the transmission video storage D / A conversion unit 106, and a call control signal input from the CPU 100. The control signal such as is subjected to frequency multiplexing processing to generate a transmission signal. The signal transmitting / receiving unit 107 transmits the transmission signal to the parent device 2 via the transmission path 3.

  In the configuration of FIG. 2, an image captured by the camera unit 11 at a cycle of 30 screens / second is a cycle according to the first cycle (7.5 screens / second), and the visitor determination unit 103 determines whether there is a visitor. To be judged. Also, the video is read from the transmission video storage D / A conversion unit 106 at a cycle according to the second cycle (30 screens / second) and transmitted to the parent device 2. Thus, since the first cycle and the second cycle are not synchronized, the second faster video from the camera unit 11 without depending on the first cycle in the video recognition unit 102 with a slow processing speed. The effect that it can send to the transmission line 3 with the period of is obtained.

  Referring to FIG. 3, base unit 2A included in the TV intercom apparatus according to the present embodiment includes a CPU 200 that controls the entire base unit 2, a liquid crystal display unit 21 that displays a visitor's video, and a visitor. A call button 22 for calling the handset 1 to start a call, an extension button 25 for calling another base unit 2 to start an extension call with another user indoors, and a visit recorded in the base unit 2 A signal transmission / reception unit 207 including a terminal block for performing an existing line communication with the playback button 24 for requesting playback of the person's video, the power supply circuit unit 28, the slave unit 1 and the other master unit 2, and a microphone (not shown) A voice call unit 23 that includes a speaker and enables voice calls with outdoor visitors and other indoor users, a video recording unit 202 that records a visitor video shot by the handset 1, and a ringing tone And a processing unit 201. That.

  The signal transmitting / receiving unit 207 receives the frequency-multiplexed transmission signal transmitted from the slave unit 1 and the master unit 2B via the transmission path 3, and converts the transmission signal into the audio signal and the video according to the control signal from the CPU 200. The control signal is separated into a signal and a paging control signal.

  The power supply circuit unit 28 generates DC 3.3V, 5V, and 12V necessary for the operation of the master unit 2 based on the AC100V that is connected to the AC100V outlet and receives power, and supplies power to each operation circuit block.

  The signal transmission / reception unit 207 is connected to the voice call unit 23, the CPU 200, the video recording unit 202, and the liquid crystal display unit 21, and the voice signal obtained by the separation is sent to the voice call unit 23, and the control signal is sent to the CPU 200. The video signal is sent to the video recording unit 202 and the liquid crystal display unit 21.

CPU 200 detects a call start signal input from signal transmission / reception unit 207 and detects the presence of a visitor near handset 1.
Further, CPU 200 detects a call end signal, which is a control signal from slave unit 1 input from signal transmission / reception unit 207, and detects that a response operation to the call operation from slave unit 1 has been performed in master unit 2B. To do.
Further, the CPU 200 inputs a call signal from the call button 22 when pressed by an indoor user. When the CPU 200 detects a call signal from the call button 22, it detects that the indoor user has pressed the call button 22, and a call response which is a control signal as a control operation for the call operation from the handset 1 or the other base unit 2. The signal is sent to the signal transmission / reception unit 207.
Further, the CPU 200 inputs a calling signal from the extension button 25 when pressed by an indoor user. When the CPU 200 detects a call signal from the extension button 25, the CPU 200 detects that an indoor user has pressed the extension button 25, and sends an extension call signal, which is a control signal, as a control operation for the call operation to the child device 1 or the parent device 2B. The data is transmitted to the signal transmission / reception unit 207.
Further, the CPU 200 inputs a reproduction request signal from the reproduction button 24 when pressed by an indoor user. When the CPU 200 detects a playback request signal from the playback button 24, the CPU 200 detects that the indoor user has pressed the playback button 24, and is recorded in the video recording unit 202 as a control operation for the playback request operation of the recorded video to the video recording unit 202. The latest visitor video is sent to the liquid crystal display unit 21.

  The video recording unit 202 is composed of a flash memory having a capacity capable of recording two videos for 30 seconds / cycle with a period of 30 screens / second, and has a ring memory structure in which new videos are overwritten and stored in order from old videos. The video signal input from the signal transmission / reception unit 207 is sequentially stored in accordance with the control signal from the CPU 200.

  The liquid crystal display unit 21 receives and displays a video signal sent from the signal transmission / reception unit 207 or the video recording unit 202 in accordance with a control signal from the CPU 200.

  The voice call unit 23 is connected to the signal transmission / reception unit 207, and outputs a voice signal input from the signal transmission / reception unit 207 to an indoor user from a speaker according to a control signal from the CPU 200. A sound to be emitted is acquired, and a sound signal based on the acquired sound is transmitted to the signal transmission / reception unit 207.

  The signal transmission / reception unit 207 performs a frequency multiplexing process on the voice signal input from the voice call unit 23 and the control signal input from the CPU 200 in accordance with the control signal from the CPU 200 to generate a transmission signal. Then, the transmission signal is transmitted to the slave unit 1 and the master unit 2B via the transmission path 3.

  Below, operation | movement of each apparatus is demonstrated based on the flowchart shown in FIGS. The processes according to these flowcharts are stored in advance as programs in a ROM (Read Only Memory) built in the CPU 100 of the slave unit 1 and a ROM built in the CPU 200 of the indoor unit 2 and read by the CPUs 100 and 200, respectively. This function is realized by being executed.

  Next, the processing of the slave unit 1 when the master unit 2A of the television intercom apparatus according to the present embodiment responds to a call of the slave unit 1 will be described using the flowchart of FIG. The process shown in the flowchart of FIG. 4 is realized mainly by the CPU 100 of the slave unit 1 reading and executing a program stored in a storage device (not shown) and controlling each unit shown in FIG.

  Referring to FIG. 4, in an initial state where the power of the TV intercom apparatus is first turned on, handset 1A executes an initialization process such as erasure of buffers in the respective internal parts (S101).

  Information of the temperature measurement unit 16, the heat source movement detection unit 14, and the month / day / time storage unit 105 used for calculation in the video recognition unit 102 and the visitor determination unit 103 is acquired at a frequency of 7.5 times per second. The CPU 100 is always stored as the latest information in the temporary storage internal register. In addition, the camera unit 11 of the slave unit 1A always performs an imaging operation while power is supplied (S102). The video recognition unit 102 analyzes the video signal based on the video captured by the camera unit 11 of the slave unit 1 at a frequency of 7.5 times per second (S103), and the visitor determination unit 103 similarly performs the analysis. In addition, a process for determining whether there is a visitor at a frequency of 7.5 times per second (S104) and, in the case of NO in S104, a process for determining whether the call button 12 has been pressed by the visitor is further performed. (S105). S103 of the process of analyzing the camera video will be described later with reference to FIG. Further, S104 of the process for recognizing the visitor will be described later with reference to FIG.

  When the visitor determination unit 103 determines that there is a visitor (YES in S104), based on that determination, or when the call button 12 of the handset 1 is pressed by the visitor (in S105) YES), based on the operation signal input from the call button 12, the CPU 100 of the child device 1 detects that there was a visitor. The CPU 100 of the child device 1 transmits a call start signal, which is a control signal, to the parent devices 2A and 2B via the signal transmission / reception unit 107 (S106).

  At the same time, when the captured image, which is one of the imaging conditions of the camera unit 11, is dark, when the amplification index represented by 255 levels indicating the degree of signal amplification to obtain an optimally bright image is 224 or more, It is determined that the captured image is dark, and lighting control of the LED 15 for illumination is performed (S107).

Further, after the call start signal is transmitted, the camera unit 11 of the slave unit 1 continues until the CPU 100 receives a response via the signal transmission / reception unit 107, with 1 minute counted by the built-in timer of the CPU 100 as an upper limit (NO in S108). The captured video is transmitted to the master unit 2A and the master unit 2B through the transmission video storage D / A conversion unit 106 and the signal transmission / reception unit 107 (S110).
If no response to the call start signal can be received from any parent device within one minute (YES in S108 for time-up), the CPU 100 regards that the visitor is absent and displays all the visitor's parent devices. When the transmission of the addressed video is interrupted and the LED illumination is turned on at the same time, a light-off process is performed (S115).
The CPU 100 acquires the movement history of the coordinates of the face image at a frequency of 7.5 times per second by the video recognition unit 102 until the time is up or the response of the parent device 2 is accumulated and stored in the analysis result accumulation unit 104 (S109).

  Assume that CPU 100 determines that the control signal received by signal transmitter / receiver 107 has received a call response signal response from parent device 2A (the call button on the parent device has been pressed) (YES in S111). In that case, bidirectional audio signal transmission / reception is performed between the slave unit 1 and the master unit 2A via the signal transmission / reception unit 107, and the voice of the resident in the house is output to the visitor from the speaker. (S112). As a result, a call using the slave unit 1 and the master unit 2A can be performed between the visitor and the person at home.

When the call with the visitor ends (YES in S113), the transmission of video and audio to the main unit 2A is stopped, and at the same time, if the LED illumination is turned on, a turn-off process is performed (S114). A series of processing in the case where there is is completed. The end of the call is detected, for example, when the call button 22 is operated on the base unit 2A.
Note that the processing in FIG. 4 is repeated until power feeding through the transmission path 3 between the slave unit 1 and the master unit 2A is interrupted.

  Next, processing of base unit 2A when base unit 2A of the television intercom apparatus according to the present embodiment responds to a call of slave unit 1 will be described using the flowchart of FIG. The process shown in the flowchart of FIG. 5 is realized mainly by reading and executing a program stored in a storage device (not shown) of CPU 200 of base unit 2A, and controlling each unit shown in FIG.

Referring to FIG. 5, in the initial state when the TV intercom apparatus is first turned on, base unit 2A executes initialization processing such as erasing the buffers of the respective internal parts and starts feeding power to slave unit 1 through transmission path 3. (S201).
In step S201, after the initialization power supply process in the initial state in which the power of the TV intercom apparatus is turned on is completed, the CPU 200 waits until a recorded video reproduction operation signal is detected (YES in S202) or the master unit 2B. Until an extension request operation signal is detected (YES in S203) or until the presence of a visitor in the vicinity of handset 1 is detected and a call start signal is detected (YES in S204), the operation standby state continues. .

  In S202, when the playback button 24 of the master unit 2A is pressed by a resident in the house, the CPU 200 detects that there is a playback request for the recorded video based on the operation signal input from the playback button 24 (YES in S202). ). When the CPU 200 detects the playback request signal from the playback button 24, the CPU 200 detects that the indoor user has pressed the playback button 24, and the video recording unit 202 displays the latest visitor video recorded in the video recording unit 202. Until the 10-second playback recorded in (1) is completed (YES in S216), the data is sent to and displayed on the liquid crystal display unit 21 (S215).

  When the indoor user presses the extension button 25 in S203, the CPU 200 detects an extension request operation signal from the extension button 25 (YES in S203), and sends a call start signal to the master unit 2B via the signal transmission / reception unit 207. Then, it is transmitted (S217). Further, after the call start signal is transmitted, the CPU 200 receives a response via the signal transmission / reception unit 207 (YES in S219), with the upper limit being one minute counted by the built-in timer of the CPU 200 (YES in S218). The voice signal is transmitted and received, and a call using the master unit 2A and the master unit 2B becomes possible (S220). When the call using the parent device 2A and the parent device 2B is completed (YES in S221), the operation standby state continues again.

  In S204, the CPU 200 detects the presence of a visitor in the vicinity of the child device 1 by receiving a call start signal from the child device 1 via the signal transmission / reception unit 207 (YES in S204). When the call start signal is detected, the CPU 200 controls the ring tone processing unit 201. Thereby, the ringing tone generated by the ringing tone processing unit 201 is output from the speaker constituting the base unit voice call unit 23 (S205).

  At the same time, since the video signals photographed by the camera unit 11 of the slave unit 1 are sequentially transmitted via the signal transmission / reception unit 207, the CPU 200 receives this and displays the video of the visitor on the liquid crystal display unit 21. In addition to the display, the recording control of the visitor video displayed on the liquid crystal display unit 21 to the video recording unit 202 is performed for 10 seconds (S206). As a result, since the visitor's video is recorded even when the resident is absent, the visitor's video can be confirmed by performing the reproduction operation.

Here, when CPU 200 detects that call button 22 of parent device 2A has been pressed by a home resident (YES in S209), CPU 200 transmits a call response signal from signal transmission / reception unit 207 to child device 1 (S210). . From the handset 1 that has received the response signal, it is possible to continuously receive the video of the visitor (S211).
Here, when the home person first responds to the visitor by transmitting a call response signal from the base unit 2B (YES in S207), the CPU 200 sends a ringing tone and sends it to the liquid crystal display unit 21. The video display is terminated, and the control shifts to a standby state (S214).
If the call button 22 is not pressed by the home person within 55 seconds, a time-up determination is made (YES in S208), the CPU 200 terminates the sending of the ringing tone and the video display on the liquid crystal display unit 21, and the control is performed. The process shifts to a standby state (S214).

  In step S212, bidirectional audio signals are transmitted / received between the child device 1 and the parent device 2A, and the voice of the visitor is output from the speaker. As a result, a call using the handset 1 and the base unit 2A can be performed between the visitor and the person at home.

When the call with the visitor is ended by an end-of-call operation such as pressing the call button 22 again in step S213 (YES in S213), transmission / reception of audio to / from the slave unit 1 and reception of video are stopped in step S214. The display on the call liquid crystal display unit 21 is also stopped, and the call processing with the visitor ends.
The process of FIG. 5 is repeated until the AC 100V plug, which is the power source, is removed and the power supply of AC 100V to the power supply circuit unit 28 of the indoor parent device 2A is stopped.
Although the processing of FIG. 5 has been described for the master unit 2A, the same applies to the master unit 2B.

  Next, details of processing in which the handset 1 of the television intercom apparatus according to the present embodiment analyzes a camera video will be described using the flowchart of FIG.

  When analyzing the video shot by the camera unit 11 of the slave unit 1, the date information counted by the date / time storage unit 105 is past 14:00 in August, and the shooting conditions of the camera unit 11 are in 255 levels. When the signal amplification degree is low as 100 (NO in S301, YES in S303), the camera image is a very good condition of S / N suitable for image analysis in clear sky in midsummer. Under the condition that there is a high possibility that there is a large amount of exposure, analysis control in the CPU 100 is performed (S304). On the other hand, when the time is 18:00 and the signal amplification degree expressed by the 255 stage of the photographing condition of the camera unit 11 is as high as 230 (YES in S301), the camera video is not suitable for image analysis in the dark night of street lamp illumination. Since / N is a bad condition, analysis control is performed by the CPU 100 in a state where the determination threshold value for determining the moving body is large in order to prevent erroneous determination (S302). Note that the imaging condition acquisition unit is a unit that acquires an amplification index represented by 255 levels, and exists in the camera unit 11. The date / time information storage means is a means for storing the season (month / day) and time obtained by the time measurement means as the month / date / time storage unit 105.

  Next, the CPU 100 of the slave unit 1 compares the difference in luminance information between the two images captured by the two camera units 11 having a time difference of 1 / 7.5 second accumulated in the recognition image accumulation unit 101. An operation is performed (S305). If there is a change in the size exceeding the determination threshold adjusted in step S302 due to a visitor, a passerby, or a shake of a plant in the captured video, the amount of movement and the screen of the moving object are calculated by this calculation. The internal coordinates are set in the CPU 100 as internal temporary storage information. Also, the luminance distribution in the video captured by the camera unit 11 stored in the recognition video storage unit 101 is analyzed in pixel units, and the average value of the entire luminance information screen represented by 255 levels is 180. Thus, the presence / absence of a group of pixels brighter than 30% and the high luminance part coordinate value are set in the CPU 100 as internal temporary storage information (S306).

  If the coordinate information of the moving object obtained as a result of the above calculation and the high-luminance part coordinate information do not match, it is temporarily stored in the CPU 100 that an artificial moving object that is not a natural phenomenon exists in the imaging area. Information is set (YES in S307, S308).

Next, the CPU 100 is limited to the image recognition target area indicated by the analysis result accumulation unit 104 in which the detection history of visits of past visitors is accumulated in the image area photographed by the camera unit 11. Analyzes color information. Judgment that a human face exists when there is a collection of skin-colored pixels below the dark-colored pixels analogized as hair or the like, and the proportion of the skin-colored pixels in the video area photographed by the camera unit 11 is 10% or more The coordinates are set as internal temporary storage information in the CPU 100 (YES in S310, S311). If the summer flag is set in S304 at the time of analysis in S310, an area that exceeds 10% if the ratio of the skin color pixel to the image area photographed by the camera unit 11 is not 30% or more. By determining that the skin color area located at the top of the screen is a human face, a process of reducing the possibility of erroneously recognizing the exposed skin color arm is performed.
By performing the processing of the above steps, any of the three types of non-detection, moving object detection, and face detection is set as a recognition result flag that is internal temporary storage information of the CPU 100.

  Next, details of processing in which the handset 1 of the television intercom apparatus according to the present embodiment determines whether there is a visitor will be described using the flowchart of FIG.

When the detection result of the heat source movement detection unit 14 of the child device 1 is a determination that there is a moving body (YES in S401), the CPU 100 determines that the video area is imaged by the camera unit 11 when the detection result flag is face detection. In addition, it can be determined that there is a human visitor within a distance of 10 m from the handset 1 (YES in S405).
Even when the face is not detected by the image recognition unit 102 due to the visitor facing away from the child device 1 or the like, if the detection result flag is a moving body detection, infrared rays are similarly emitted. It can be determined that there is a visitor who is a calling person (NO in S405, YES in S406).
However, since the pyroelectric sensor used in the heat source movement detection unit 14 requires a temperature difference of 4 ° C. or more from the background, when the ambient temperature measured by the temperature measurement unit 16 is 31 ° C. or more, the heat source The determination result of the movement detection unit 14 cannot be trusted. Therefore, regardless of the detection result of the heat source movement detection unit 14, if the detection result flag is face detection, the CPU 100 performs a determination process when a visitor is present in the video area captured by the camera unit 11 (S402). YES, YES in S404).

In addition, when the visitor walks in a straight line toward the child device 1 or stops, the heat source movement detection unit 14 erroneously determines “no heat source moving object”. Regardless of the detection result of the heat source movement detection unit 14, if the detection result flag is face detection, the CPU 100 determines that a visitor is present in the video area being shot by the camera unit 11 (NO in S 401, NO in S402, YES in S404).
However, if the detection result of the heat source movement detection unit 14 is a determination that there is no heat source movement and the detection result flag is a moving object, a moving object of a non-human artifact is an erroneous recognition in the image recognition unit 102, and YES in step S403 The CPU 100 determines that “there is no visitor” (S408).

The configuration of the television intercom apparatus according to the present invention is not limited to the above-described configuration, and the master units 2A and 2B only have to be configured so that at least video transmission / reception can be performed with the slave unit 1.
Further, in the present embodiment, the digital signal video is converted into an analog signal and modulated / demodulated by the signal transmission / reception unit 107. However, the signal transmission / reception unit 107 may transmit the digital signal as it is and perform modulation / demodulation by the signal transmission / reception unit 107.
Furthermore, the recording medium for recording the captured video is not limited to the flash ROM built in the master unit 2, but can be exchanged like an SD memory card (registered trademark) that allows the user to select a storage medium having a capacity according to the application. A possible (detachable) structure may be adopted.

  The embodiment disclosed this time should be considered as illustrative in all points and not restrictive. The scope of the present invention is defined by the terms of the claims, rather than the description above, and is intended to include any modifications within the scope and meaning equivalent to the terms of the claims.

It is a block diagram of the television intercom apparatus which concerns on this Embodiment. It is a figure which shows the specific example of a structure of the intercom handset. It is a figure which shows the specific example of a structure of the intercom main unit. 6 is a flowchart when a call to the interphone master unit 2 occurs in the interphone slave unit 1; It is a flowchart which shows the process which responds to the call of the intercom handset 1 by the intercom base unit 2. It is a flowchart which shows the detail of the analysis process S103 of the camera image | video of the intercom cordless handset 1. It is a flowchart which shows the detail of the visitor recognition judgment process S104 of the intercom handset 1.

Explanation of symbols

DESCRIPTION OF SYMBOLS 1 Interphone subunit | mobile_unit 2,2A, 2B Interphone main | base station 3 Communication transmission path, 11 Camera part 12 Call button 13,23 Voice call part 14 Heat source movement detection part 15 LED illumination part 16 Temperature measurement part 21 Liquid crystal display part 22 Call button 24 Recording / playback button 25 Extension call button 100, 200 CPU 101 Recognition video storage unit 102 Video recognition unit 103 Analysis result storage unit 104 Month / day time storage unit 105 Transmission video storage D / A conversion unit 107, 207 Signal transmission / reception unit 108, 208 Power Supply Circuit Unit 109 Calling Mode Storage Unit 201 Ringing Sound Processing Unit 202 Video Recording Unit

Claims (7)

In a television intercom apparatus comprising a slave unit having a camera unit provided at a position where a visitor can be photographed, and a plurality of master units capable of communicating with the slave unit and having a display unit,
The slave unit is further
Video recognition means for inputting a video imaged and output by the camera unit according to a first cycle, and analyzing the presence or absence of a visitor based on the input video image;
Visitor determination means for determining presence or absence of a visitor based on the video recognition result performed by the video recognition means;
Video transmitted and captured by the camera unit is input in accordance with a second cycle, and the input video is transmitted to the master unit, and
The master unit is further
Receiving video display means for receiving the video transmitted from the video transmission means and displaying the video on the display unit;
TV intercom device consisting of The video output by the camera unit is a video shot with a first pixel configuration,
The video recognizing unit converts the video from the camera unit into a second pixel configuration and inputs it, and analyzes the presence / absence of a visitor based on the input video information,
The video transmission means is configured such that the video from the camera unit is converted into a third pixel configuration and input, and the input video is transmitted to the parent device.
The television intercom apparatus according to claim 1, comprising: The slave is
Analysis result storage means for storing the analysis result in the video recognition means for a certain period,
Based on the accumulated information of the analysis result accumulating means, an analysis area for analyzing presence / absence of a visitor is set in the video image area captured by the camera unit input to the video recognition means,
The television intercom apparatus according to claim 1 or 2, comprising the above. The slave unit is further
It has a means for acquiring video shooting conditions in the camera unit,
Based on video shooting conditions in the camera unit obtained under the shooting conditions, analysis of presence or absence of visitors in the video image area taken by the camera unit input to the video recognition means,
The television intercom apparatus according to any one of claims 1 to 3, further comprising: The slave unit is further
Having time measuring means,
Based on the date and time information obtained by the time measuring means, the presence or absence of visitors is analyzed in the video image area taken by the camera unit input to the video recognition means.
The television intercom apparatus according to claim 1, comprising: The slave unit is further
Heat source movement detection means provided at a position where infrared rays generated from a visitor can be detected, the heat source detection result of the heat source movement detection means, the video recognition result of the video recognition means, and the imaging condition acquisition means Based on the acquired video imaging conditions in the camera unit, the visitor determination means determines the presence or absence of a visitor,
The television intercom apparatus according to claim 1, comprising: The slave unit is further
Having a temperature measuring means for measuring the temperature,
Based on the temperature information of the temperature measurement means, the heat source detection result of the heat source movement detection unit, and the video recognition result of the video recognition means, the visitor determination means determines whether or not there is a visitor.
The television intercom apparatus according to any one of claims 1 to 6, further comprising: