JP2019115052A - Head detector and head detection method and monitored person monitoring device - Google Patents

Abstract

To detect the head of a monitored person more precisely.SOLUTION: A sensor unit SU includes a camera 12 for acquiring an image capturing a monitoring object, i.e., a monitored person, from above the monitored person, a circular area extraction part 322 for extracting a substantially circular area, as the circular area, on the basis of the image captured by means of the camera 12, a low index of reflection area extraction part 323 for extracting an area of a predetermined index of reflection or less for infrared radiation, as the low index of reflection area, on the basis of the image captured by means of the camera 12, a marker detector 325 for detecting a predetermined marker on the basis of the image captured by means of the camera 12, and a head selection part 324 for detecting the head. The head selection part 324 selects the area of head on the basis of residual circular area and low index of reflection area excepting the area having the marker from the circular area and the low index of reflection area, and detects the head, and the marker is identifiable only with infrared light, and the index of reflection for infrared light exceeds the predetermined index of reflection.SELECTED DRAWING: Figure 2

Description

  The present invention relates to a head detection device and a head detection method for detecting the head of a monitored person to be monitored. And this invention relates to the to-be-monitored person monitoring apparatus using this head detection apparatus.

  Japan's aging society, which is the ratio of the population over the age of 65 to the total population, is due to the improvement of living standards following the postwar high economic growth, the improvement of the sanitation environment and the improvement of medical care. Has become a super-aging society with over 21%. Also, in 2005, the total population was about 127.65 million, compared with about 25.56 million for the elderly population aged 65 and over, in 2020 the total population was about 124.11 million. The elderly population is expected to be about 34.56 million. Such an aging society is expected to increase compared to the number of nurses who need nursing or nursing care due to illness, injury or old age, which occurs in a normal society that is not an aging society. And Japan is also a low birthrate society, for example, with a total fertility rate of 1.43 in 2013. Therefore, old-age care for elderly people (spouses, children, brothers) taking care of old people who need nursing or care has also been awake.

  Those who need nursing care and those who need nursing care enter facilities such as hospitals and welfare facilities for the elderly (in Japan, the facilities for entering the elderly for the elderly, nursing homes for the elderly and special nursing homes for the elderly, etc.) and receive their nursing and nursing care. In such a facility, there is a problem that a nurse who needs care is injured, for example, by falling from a bed or falling while walking, or he gets out of bed and hesitates. It is necessary to respond to such a situation as quickly as possible, and if such a situation is left as it is, there is a possibility that it will develop into a bigger problem. And carers, etc., have confirmed their safety and condition by patrolling regularly.

  However, the number of nurses etc. can not keep up with the number of nurses etc., and the nursing industry and the care industry are chronically understaffed. Furthermore, since the number of nurses and carers etc. decreases during the night shift and night shift hours compared with the day shift hours, the workload per person increases, so it is required to reduce the load. In addition, the elderly care is not an exception at the facilities, and it is often seen that an elderly nurse or the like cares for an elderly nurse or the like. Generally, when you get older, your physical strength declines, and even if you are healthy, the burden of nursing etc. becomes heavier than younger nurses and so on, and its movement and judgment also become slower.

  In order to reduce such labor shortages and the burden of nurses etc., there is a need for techniques that complement nursing work and care work. For this reason, in recent years, a monitored person monitoring device for monitoring (monitoring) a monitored person to be monitored such as a nurse who needs to be monitored has been researched and developed.

  One such technique is, for example, a fall detection device disclosed in Patent Document 1. The fall detection device disclosed in Patent Document 1 includes a camera, an image processing unit that takes in an image signal output from the camera, and performs difference processing on the captured image, and a difference region obtained by the difference processing. And judging means for detecting a falling motion of the human body based on the area. More specifically, the determination means determines the presence or absence of a fall by the change in area per unit time in the human body.

  On the other hand, in terms of safety confirmation, a single person living alone is also the same as the nurse who needs to be a person to be monitored and is a monitored person.

JP, 2000-207665, A

  By the way, since the size of the area of the human body shown in the image changes depending on the presence or absence of the fall, the fall detection device disclosed in Patent Document 1 detects the fall by the area change per unit time in the human body area. Since the size of the head area shown in the image also changes depending on the presence or absence of a fall, a method for detecting a fall based on the size of the head area can be considered. In this method, in the image captured from above, the head appears in a substantially circular shape, so a circular area is detected from the image as the head area. However, since a substantially circular object other than the head such as a pillow, a basin, a ball, a trash can and a futon handle is also included in the actual image, the area of these objects is erroneously detected as a head area, A fall may be misdetected.

  On the other hand, if these objects exist at a fixed position, the false detection can be avoided by previously excluding the regions of the objects at the fixed position from the region for detecting the head. However, these objects do not always exist at fixed positions.

  The present invention is an invention made in view of the above-mentioned circumstances, and an object thereof is a head detection device and a head capable of detecting the head of a person to be monitored with high accuracy based on an image taken from above the person to be monitored It is an object of the present invention to provide a part detection method and a monitored person monitoring device using the head detection device.

  As a result of various studies, the present inventor has found that the above object can be achieved by the present invention described below. That is, in the head detection device according to one aspect of the present invention, an image acquisition unit that acquires an image obtained by capturing an image of a person to be monitored from above the person to be monitored, and an image acquired by the image acquisition unit And a circular area extraction unit that extracts a substantially circular area as a circular area, and an area having a predetermined reflectance or less for infrared light based on the image acquired by the image acquisition unit. A low reflectance area extraction unit extracted as a low reflectance area, a marker detection unit detecting a predetermined marker based on the image acquired by the image acquisition unit, and a circular shape extracted by the circular area extraction unit A head selecting unit for selecting the head region based on the region, the low reflectance region extracted by the low reflectance region extraction unit, and the marker detected by the marker detection unit; The head selection unit may The circular shaped area remaining after removing the area having the marker detected by the marker detection unit from the circular shaped area extracted by the shape area extracting unit and the low reflectance area extracted by the low reflectance area extracting unit The head area is selected based on the index area to detect the head, and the marker is distinguishable only by the infrared light, and the reflectance for the infrared light is the predetermined reflectance. It is characterized by exceeding. Preferably, in the head detection apparatus described above, since the probability of being able to image the entire monitored person is high, the image is scheduled when the head of the monitored person is located in the bedclothes where the monitored person is lying down The image is taken from directly above the preset head planned position.

  As a result of various studies, the inventor has found that the head has a relatively low reflectance to infrared light (infrared light). From this finding, the head detection device detects the head in consideration of not only the substantially circular shape but also the reflectance to infrared light, so the head of the person to be monitored is detected based on the image of the person to be monitored from above It is possible to detect a part more accurately. Further, since the marker is also considered, the head of the person to be monitored can be detected more accurately. Further, by adding a marker to a head approximating a head other than the head, the head detection device excludes the circular area having the marker and the low reflectance area, thereby forming the circular area. The non-head portion can be removed from the circular shaped region extracted by the extraction unit and the low reflectance region extracted by the low reflectance region extraction unit, and the head of the person to be monitored can be detected more accurately. Moreover, the misdetection in a marker detection part can be reduced by using the marker which can be identified only by infrared light. Since the marker can not be viewed, the head detection device does not give a sense of discomfort to the monitored person without being aware of the presence of the marker when mounted on the monitored person monitoring device.

  Further, in another aspect, in the above-described head detection device, the low reflectance area extraction unit performs low reflection on a region having a predetermined reflectance or less from a circular region extracted by the circular region extraction unit. It is characterized in that it is extracted as a rate area.

  Such a head detection device narrows the circular area extracted by the circular area extraction unit as a candidate for the head area (head area), and considers the reflectance to infrared light from the narrowed candidates Since the head is detected, it is possible to reduce erroneous detection of an object having a low reflectance as compared with the case where a low reflectance area is extracted from the entire image acquired by the image acquisition unit.

  Further, in another aspect, in the head detection device described above, the circular area extraction unit is configured to generate a substantially circular area from the low reflectance area extracted by the low reflectance area extraction section. It is characterized by being extracted as

  Such a head detection device narrows down the low reflectance area extracted by the low reflectance area extraction unit as a head area candidate, and detects the head from the narrowed down candidates in consideration of the substantially circular shape. The amount of information processing can be reduced compared to the case where a substantially circular circular area is extracted from the entire image acquired by the image acquisition unit.

  In another aspect, in the above-described head detection device, the marker is a marker having a specific shape.

  Such a head detection device can reduce false detection in the marker detection unit by using a marker having a specific shape.

  In the head detection method according to another aspect of the present invention, an image acquisition step of acquiring an image obtained by capturing an image of a person to be monitored from above the person to be monitored, and the image acquisition step Based on the extracted image, a circular area extraction step of extracting a substantially circular area as a circular area, and a predetermined reflectance for infrared light based on the image acquired in the image acquisition step An area is extracted as a low reflectance area, a low reflectance area extraction process, a circular shaped area extracted in the circular shape area extraction process, and a low reflectance area extracted in the low reflectance area extraction process A head selecting step of selecting an area and detecting the head, and a marker detecting step of detecting a predetermined marker based on the image acquired in the image acquiring step, the head selecting step Is the circular area The remaining circular shape area and the low reflectance area except the circular shape area extracted in the extraction process and the low reflectance area extracted in the low reflectance area extraction process excluding the area having the marker detected in the marker detection process Selecting the area of the head based on the detection of the head, and the marker is distinguishable only by the infrared light, and the reflectance for the infrared light exceeds the predetermined reflectance It is characterized by

  Since such a head detection method detects the head in consideration of not only the substantially circular shape but also the reflectance to infrared light, the head of the person to be monitored is detected based on an image of the person to be monitored from above. It can detect more accurately.

  Furthermore, the to-be-monitored person monitoring device according to another aspect of the present invention is a head detection unit that detects a head of a to-be-monitored person to be monitored from an image captured from above the to-be-monitored person; An action detection unit for detecting a predetermined action of the person to be monitored based on the head of the person to be monitored detected by a head detection unit, and a notification for notifying the outside of the predetermined action detected by the action detection unit And the head detection unit includes any of the above-described head detection devices.

  Such a person-to-be-monitored monitoring apparatus detects the head by using any of the above-described head detection devices, so that the head of the person to be monitored can be made more accurate based on an image of the person to be monitored It is possible to detect well, and as a result, it is possible to detect predetermined behavior of the person to be monitored more accurately.

  The head detection device, the head detection method, and the monitored person monitoring device according to the present invention can more accurately detect the head of the monitored person based on an image obtained by imaging the monitored person from above.

It is a figure which shows the structure of the to-be-monitored person monitoring system in embodiment. It is a figure which shows the structure of the sensor apparatus in the to-be-monitored person monitoring system of 1st Embodiment. It is a figure which shows the mode of arrangement | positioning of the sensor apparatus in the to-be-monitored person monitoring system of embodiment. It is a flowchart which shows operation | movement of the 1st aspect of the sensor apparatus in the to-be-monitored person monitoring system of 1st Embodiment. It is a figure for demonstrating an example of the image acquired by the sensor apparatus in the to-be-monitored person monitoring system of 1st Embodiment. It is a figure which shows the pattern of a futon. It is a flowchart which shows operation | movement of the 2nd aspect of the sensor apparatus in the to-be-monitored person monitoring system of 1st Embodiment. It is a figure which shows the structure of the sensor apparatus in the to-be-monitored person monitoring system of 2nd Embodiment. It is a flowchart which shows operation | movement of the 1st aspect of the sensor apparatus in the to-be-monitored person monitoring system of 2nd Embodiment. It is a flowchart which shows operation | movement of the 2nd aspect of the sensor apparatus in the to-be-monitored person monitoring system of 2nd Embodiment.

  Hereinafter, an embodiment according to the present invention will be described based on the drawings. In addition, the structure which attached | subjected the same code | symbol in each figure shows that it is the same structure, and abbreviate | omits the description suitably. In the present specification, suffixes are generally indicated by reference numerals without suffixes, and individual configurations are indicated by subscripts.

  The monitored person monitoring apparatus comprises an activity detecting unit for detecting a predetermined activity set in advance in the monitored person to be monitored, and a notifying unit for notifying the predetermined activity detected by the activity detecting unit to the outside And. And the said to-be-monitored person monitoring apparatus is further provided with the head detection part which detects a head from the image which imaged the said to-be-monitored person from the upper direction of the said to-be-monitored person. The predetermined behavior of the person to be monitored is detected based on the head of the person to be monitored detected by the head detection unit. Such a person-to-be-monitored monitoring apparatus may be realized integrally as one device, or may be realized as a system with a plurality of devices. And when a to-be-monitored person monitoring apparatus is implement | achieved by several apparatuses, the said action detection part may be mounted in any of these several apparatuses. As an example, an embodiment of the monitored person monitoring device will be described here in a case where the monitored person monitoring device is realized by a plurality of devices as a system. In addition, also when the said to-be-monitored person monitoring apparatus is comprised integrally as one apparatus, and implement | achieved, the said to-be-monitored person monitoring apparatus can be comprised similarly to the following description. Also, as an example, the case where the action detection unit is mounted on the sensor device SU described later together with the head detection unit and the notification unit will be described here, but other devices in the system, for example, Also in the case where the management server SV, the fixed terminal SP or the mobile terminal TA is mounted, the above-mentioned person-to-be-monitored monitoring apparatus can be configured as in the following description.

First Embodiment
FIG. 1 is a diagram showing a configuration of a monitored person monitoring system in the embodiment. FIG. 2 is a diagram showing a configuration of a sensor device in the monitored person monitoring system of the first embodiment. FIG. 3 is a view showing the arrangement of the sensor device in the monitored person monitoring system of the embodiment.

  The monitored person monitoring system MSa in the first embodiment of the example in which the monitored person monitoring device is realized as a system is a monitored person (observed person) who should be monitored (should watch) The monitoring target person Ob is detected by detecting a predetermined action set in advance in Ob (Ob-1 to Ob-4) and, for example, as shown in FIG. 1, one or a plurality of sensor devices SUa (SUa-1 to SUa-4), management server device SV, fixed terminal device SP, and one or more portable terminal devices TA (TA-1 and TA-2) It is wirelessly connected communicably via a network (network, communication line) NW such as a LAN (Local Area Network), a telephone network, and a data communication network. The network NW may be equipped with relays such as repeaters, bridges, routers, and cross connects that relay communication signals. In the example illustrated in FIG. 1, the plurality of sensor devices SUa-1 to SUa-4, the management server device SV, the fixed terminal device SP, and the plurality of portable terminal devices TA-1 and TA-2 include radios including the access point AP. They are communicably connected to each other by a LAN (for example, a LAN according to the IEEE 802.11 standard) NW.

  The monitored person monitoring system MSa is disposed at an appropriate place according to the monitored person Ob. The monitored person (watching target person) Ob is, for example, a person who needs nursing due to illness or injury, a person who needs care due to a decrease in physical ability or the like, or a single person living alone. In particular, from the viewpoint of enabling early detection and coping, the person to be monitored Ob is a person who needs the detection when a predetermined adverse event such as an abnormal condition occurs in the person. preferable. For this reason, the person-to-be-monitored monitoring system MSa is suitably disposed in buildings such as hospitals, welfare facilities for the elderly and dwelling units according to the type of the person to be monitored Ob. In the example illustrated in FIG. 1, the monitored person monitoring system MS is disposed in a building of a care facility provided with a plurality of living rooms RM in which a plurality of monitored persons Ob reside, and a plurality of rooms such as a nurse station.

  The sensor device SUa has a communication function of communicating with the other devices SV, SP, and TA via the network NW, detects the monitored person Ob, and transmits the detection result to the management server device SV. For example, as shown in FIG. 2, such a sensor device SUa includes a sensor unit 1, a sound input / output unit 2, a control processing unit 3, a communication interface unit (communication IF unit) 4, and a storage unit 5. Equipped with

  The sensor unit 1 is a device that is connected to the control processing unit 3 and detects predetermined amounts of predetermined in the monitored person Ob under the control of the control processing unit 3. The predetermined amounts are appropriately determined in accordance with predetermined actions to be detected in the monitored person Ob. In the present embodiment, the predetermined action is the rising, leaving, falling, and movement abnormality of the monitored person Ob. In order to detect these, the sensor unit 1 includes, for example, a Doppler sensor 11 and a camera 12.

  The Doppler sensor 11 is a body movement sensor that transmits a transmission wave, receives a reflection wave of the transmission wave reflected by an object, and outputs a Doppler signal of a Doppler frequency component based on the transmission wave and the reflection wave. And outputting the Doppler signal to the control processing unit 3. The transmission wave may be an ultrasonic wave, a microwave or the like, but in the present embodiment, it is a microwave. The microwave can be transmitted through the clothes and reflected by the body surface of the monitored person Ob, so that the movement of the body surface can be detected even if the monitored person Ob wears clothes, which is preferable.

  The camera 12 is a device that is connected to the control processing unit 3 and generates an image (image data) according to the control of the control processing unit 3. The camera 12 is disposed so as to be able to monitor the space (location space, a room RM in the arrangement location in the example shown in FIG. 1) where the monitored subject Ob to be monitored is scheduled to be located, The image pickup target is picked up from above, an image (image data) which looks over the image pickup target is generated, and the image of the pickup target is output to the control processing unit 3.

  In the present embodiment, as shown in FIG. 3, since the probability of being able to image the whole of the monitored person Ob is high, the camera 12 is in the bedding BT such as a bed or the like lying on the monitored person Ob. The sensor device SU is disposed so that an imaging target can be imaged from directly above a preset head planned position (usually, the arrangement position of a pillow) where the head is planned to be positioned. The sensor device SU acquires a Doppler signal measured from above the monitored person Ob by the Doppler sensor 11 of the sensor unit 1, and the camera 12 of the sensor unit 1 monitors the monitored person Ob by the monitored person Ob. An image taken from the top, preferably an image taken from directly above the planned head position is obtained. As described later with reference to FIG. 5, a pillow PI is placed on the bedding BT, a duvet BD including a pattern PT is hung, and a trash can WB and a basin BA are placed in the room RM of the monitored person Ob. However, illustration of these is omitted in FIG.

  Such a camera 12 may be a device that generates an image of visible light, but in the present embodiment, it is a device that generates an image of infrared light so that the monitored person Ob can be monitored even in relatively dark conditions. is there. Such a camera 12 is disposed, for example, in the present embodiment, an imaging optical system for forming an optical image of infrared in an imaging object on a predetermined imaging surface, a light receiving surface being made to coincide with the imaging surface An image sensor that converts an infrared optical image of the imaging target into an electrical signal, and image data processing representing an infrared image of the imaging target by performing image processing on an output of the image sensor It is a digital infrared camera provided with an image processing unit and the like to be generated. In the present embodiment, the imaging optical system of the camera 12 is preferably a wide-angle optical system (a so-called wide-angle lens (including a fisheye lens)) having an angle of view capable of imaging the entire room RM provided.

  The sound input / output unit 2 is a circuit connected to the control processing unit 3 for acquiring an external sound and inputting it to the sensor unit SU, and according to the control of the control processing unit 3 according to the electric signal representing the sound It is a circuit for generating and outputting sound. The sound input / output unit 2 includes, for example, a microphone that converts acoustic vibration of sound into an electrical signal, and a speaker that converts the electrical signal of sound into acoustic vibration of sound. The sound input / output unit 2 outputs an electrical signal representing an external sound to the control processing unit 3, and converts the electrical signal input from the control processing unit 3 into acoustic vibration of sound and outputs it.

  The communication IF unit 4 is a communication circuit that is connected to the control processing unit 3 and performs communication according to the control of the control processing unit 3. The communication IF unit 4 generates a communication signal containing the data to be transferred, which is input from the control processing unit 3, according to the communication protocol used in the network NW of the monitored person monitoring system MSa, and generates the generated communication signal. Transmit to other devices SV, SP, TA via the network NW. The communication IF unit 4 receives communication signals from the other devices SV, SP, and TA via the network NW, extracts data from the received communication signals, and allows the control processing unit 3 to process the extracted data. Data to the control processing unit 3. The communication IF unit 4 further uses, for example, standards such as Bluetooth (registered trademark) standard, IrDA (Infrared Data Association) standard, USB (Universal Serial Bus) standard, etc. to input / output data to / from an external device. May be provided with an interface circuit for

  The storage unit 5 is a circuit that is connected to the control processing unit 3 and stores various predetermined programs and various predetermined data according to the control of the control processing unit 3. The various predetermined programs include, for example, a control processing program such as a monitoring processing program that executes information processing related to monitoring of the monitored person Ob. The monitoring processing program includes a head detection processing program for detecting the head of the person to be monitored Ob based on the output of the sensor unit 1 and a predetermined action on the person to be monitored Ob based on the output of the sensor unit 1. An action detection processing program, a notification processing program for notifying the predetermined action detected by the action detection program to the outside, a moving image captured by the camera 12 by streaming to the fixed terminal apparatus SP or the portable terminal apparatus TA that requested the moving image It includes a streaming processing program to be distributed, and a nurse call processing program for voice communication with the fixed terminal apparatus SP or the portable terminal apparatus TA by using the sound input / output unit 2 or the like. The head detection processing program includes a moving body extraction program for extracting a moving body area as a human body area of the monitored person Ob from an image acquired by the camera 12 of the sensor unit 1, and a moving body area extracted by the moving body extraction program. And a circular area extraction program for extracting a substantially circular area as a circular area, and a low reflectance for an area having a reflectance less than a predetermined reflectance for infrared light based on the moving body area extracted by the moving body extraction program. A head region is selected based on a low reflectance area extraction program to be extracted as an area, a circular shape area extracted by the circular shape area extraction program, and a low reflectance area extracted by the low reflectance area extraction program And a head selection program for detecting the head. The action detection processing program, as one aspect, detects a predetermined action in the monitored person Ob based on the head of the monitored person Ob detected by the head detection processing program. The various predetermined data includes data necessary for executing each of these programs, data necessary for monitoring the monitored person Ob, and the like. The storage unit 5 includes, for example, a ROM (Read Only Memory), which is a nonvolatile storage element, and an EEPROM (Electrically Erasable Programmable Read Only Memory), which is a rewritable nonvolatile storage element. The storage unit 5 includes, for example, a random access memory (RAM) serving as a working memory of the control processing unit 3 that stores data and the like generated during execution of the predetermined program.

  The control processing unit 3 controls the respective units of the sensor device SUa according to the functions of the respective units, and the sensor unit 1 detects predetermined amounts set in advance in the person to be monitored Ob, and the detected predetermined amount It is a circuit for detecting and notifying a predetermined action in the person to be monitored Ob based on various amounts. The control processing unit 3 includes, for example, a central processing unit (CPU) and peripheral circuits thereof. The control processing unit 3 executes the control processing program to execute the control unit 31, the head detection processing unit 32, the action detection processing unit 33, the notification processing unit 34, the streaming processing unit 35, and the nurse call processing unit 36. Functionally equipped.

  The control unit 31 controls each part of the sensor device SUa in accordance with the function of each part, and controls the entire control of the sensor device SU.

  The head detection processing unit 32 detects the head of the monitored person Ob based on the output of the sensor unit 1. More specifically, the head detection processing unit 32 detects the head of the monitored person Ob based on the image captured by the camera 12 of the sensor unit 1 and captured from the upper side of the monitored person Ob. To this end, the head detection processing unit 32 functionally includes a moving body extraction unit 321, a circular area extraction unit 322, a low reflectance area extraction unit 323, and a head selection unit 324.

  The moving body extraction unit 321 extracts a moving body region from the image acquired by the camera 12 of the sensor unit 1 as a human body region of the monitored person Ob by using a method such as a background difference method or an interframe difference method. In the background subtraction method, a background image is obtained in advance and stored in advance in the storage unit 5 as one of the various predetermined data, and the presence or absence of a moving object is obtained from the difference image between the image generated by the camera 12 and the background image. As a result of this determination, if there is a moving body, the area of the moving body is regarded as a moving body area. In the frame difference method, the presence or absence of a moving object is determined from the difference image of the image of the current frame and the image of the past frame (for example, the immediately preceding frame) generated by the camera 12. If there is a region of this moving body, it becomes the moving body region. The moving body extraction unit 321 notifies the extracted moving body region to the circular area extraction unit 322 in the operation of the first mode described later, and notifies the low reflectance area extraction unit 323 in the operation of the second mode described later. .

  The circular area extraction unit 322 extracts a substantially circular area as a circular area based on the image acquired by the camera 12 of the sensor unit 1. In the image captured from above, the head appears in a substantially circular shape, so extracting a circular area extracts the area (head candidate area) that is a candidate for the head. For example, the circular area extraction unit 322 extracts the circular area from the image acquired by the camera 12 of the sensor unit 1 by using the Hough transform that extracts the circular shape. Also, for example, a circular shape pattern corresponding to the shape pattern of the head is prepared in advance and stored in advance in the storage unit 5 as one of the various predetermined data, and the circular region extraction unit 322 The circular shape pattern stored in advance in 5 is searched from the image acquired by the camera 12 of the sensor unit 1 by using pattern matching, and an area matching the circular shape pattern is used as the circular area. Extract. More specifically, in the operation of the first aspect, the circular area extraction unit 322 extracts a circular area from the moving body area extracted by the moving body extraction unit 321 by, for example, the Hough transform method or the pattern matching method. The extracted circular area is notified to the low reflectance area extraction unit 323. Further, in the operation of the second aspect, the circular area extraction unit 322 performs, for example, the Hough transform method, the pattern matching method, and the like from the low reflectance area extracted by the low reflectance area extraction section 323 as described later. The circular shape area is extracted by the above, and the extracted circular shape area is notified to the head selection unit 324.

  The low reflectance area extraction unit 323 extracts an area having a predetermined reflectance or less with respect to infrared light as a low reflectance area based on the image acquired by the camera 12 of the sensor unit 1. Since the head has a relatively low reflectance for infrared light, the head candidate region is extracted by extracting the low reflectance region. The head has a low reflectance of, for example, 9%, 10% and 11% to infrared light, so 50%, preferably 30%, more preferably 10%, etc. is considered as the threshold, and low reflectance The area extraction unit 323 extracts an area equal to or less than the threshold from the image acquired by the camera 12 of the sensor unit 1 as a low reflectance area. More specifically, in the operation of the first aspect, the low reflectance area extraction unit 323 extracts, from the circular area extracted by the circular area extraction unit 322, an area equal to or less than the threshold as a low reflection area; The extracted low reflectance area is notified to the head selecting unit 324. Further, in the operation of the second aspect, the low reflectance area extraction unit 323 extracts an area equal to or less than the threshold from the moving body area extracted by the moving body extraction unit 321 as the low reflectivity area, and the extracted low reflectivity The area is notified to the circular area extraction unit 322.

  The head selecting unit 324 selects a head region based on the circular shaped region extracted by the circular shaped region extracting unit 322 and the low reflectance region extracted by the low reflectance region extracting unit 323, and the head is selected. To detect More specifically, in the operation of the first aspect, the head selection unit 324 selects the area of the head based on the low reflectance area extracted by the low reflectance area extraction unit 323 and detects the head. The action detection processing unit 33 is notified of the detected head (head area). In the present embodiment, the head selection unit 324 causes the low reflectance area extraction unit 323 to extract a low reflectance area until the number of low reflectance areas becomes one by changing the default threshold value by a predetermined value. One low reflectance area is selected as a head (head area) to detect the head. Further, in the operation of the second aspect, the head selection unit 324 selects the area of the head based on the circular area extracted by the circular area extraction unit 322, detects the head, and detects the head. The part (head area) is notified to the action detection processing part 33. In the present embodiment, the head selecting unit 324 changes the predetermined threshold value from the default threshold value by a predetermined value until the number of circular regions extracted by the circular region extracting unit 322 is one. The low reflectance area is extracted, and this one circular area is selected as a head (head area) to detect the head.

  The action detection processing unit 33 detects a predetermined action set in advance by the monitored person Ob based on the output of the sensor unit 1. In the present embodiment, as described above, the predetermined action is getting up, leaving, falling, falling or abnormal movement of the monitored subject Ob, and the action detection processing unit 33 selects the target based on the output of the sensor unit 1. Waking up, leaving, falling, and body movement abnormality in the observer Ob are detected as detection results, and the detection results are notified to the notification processing unit 34. For example, from the height (size) of the head (head area) detected by the head detection processing unit 32, the action detection processing unit 33 determines the posture (for example, standing position, sitting position, lying position, etc.) of the monitored person Ob. It judges and detects the position of this detected head, and judges the distinction between the rising, leaving and falling based on the judgment and the detected posture and position of the person to be monitored Ob. For example, when the height (size) of the head (head area) corresponds to the height (size) on the bedding BT, and the position of the head (head area) is in the bedding BT The behavior detection processing unit 33 can determine that the monitored person Ob lays down with the bedding BT, that is, is sleeping. Also, for example, when the height (size) of the head (head area) is the height (size) corresponding to a sitting position, and the position of the head (head area) is in the bedding BT The action detection processing unit 33 can determine that the monitored person Ob has got up. Also, for example, when the height (size) of the head (head area) is the height (size) corresponding to a standing position, and the position of the head (head area) is outside the bedding BT The behavior detection processing unit 33 can determine that the monitored person Ob has left the floor. Also, for example, when the height (size) of the head (head area) corresponds to the height (size) on the floor and the position of the head (head area) is outside the bedding BT The action detection processing unit 33 can determine that the monitored person Ob falls. In addition, the area of the bedding BT on the image is stored in advance in the storage unit 5 as one of the various predetermined data. Further, the height of the head (head area) is obtained from the size of the head (head area) (for example, the horizontal length, the vertical length, the area, etc.) since the image is obtained from above the monitored person Ob. be able to. Further, for example, the behavior detection processing unit 33 detects body motion of the chest (up and down movement of the chest) accompanying the breathing operation of the monitored person Ob by the Doppler sensor 11 of the sensor unit 1, and the cycle in the body motion of the chest is disturbed. When the amplitude in the body movement of the chest which is equal to or less than a preset threshold is detected, it is determined that the micro movement abnormality is present.

  The notification processing unit 34 notifies the outside of the predetermined action detected in the action detection processing unit 33 (in the present embodiment, getting up, leaving, falling, and abnormal body movement). More specifically, the notification processing unit 34 is information indicating the detected predetermined action (state, condition) (detected action information (in the present embodiment, one of wake up, leaving, fall, and body movement abnormality or Identifier information for identifying and identifying the monitored person Ob whose predetermined action has been detected)) (an identifier for identifying and identifying the sensor device SUa that is detecting the monitored person Ob) A communication signal (monitoring information communication signal) containing information) and an image etc. used for detection of the predetermined action is generated and transmitted to the management server SV by the communication IF unit 4.

  When the streaming processing unit 35 requests distribution of a moving image from the fixed terminal apparatus SP or the portable terminal apparatus TA via the network NW and the communication IF unit 4, the fixed terminal apparatus SP or the portable terminal apparatus that has made the request A moving image (for example, a live moving image) generated by the camera 12 of the sensor unit 1 is distributed to the TA by streaming reproduction via the communication IF unit 4 and the network NW.

  When the nurse call processing unit 36 receives an input operation with a nurse call push button switch (not shown) that receives a nurse call from the monitored person Ob, the fixed terminal apparatus SP or the fixed terminal apparatus SP via the communication IF unit 4 and the network NW. It enables voice communication with the portable terminal device TA.

  Four first to fourth sensor devices SUa-1 to SUa-4 are shown in FIG. 1 as an example, and the first sensor device SUa-1 is a person A who is one of the monitored persons Ob. The second sensor device SUa-2 is disposed in a living room RM-1 (not shown) of Ob-1, and is disposed in a living room RM-2 (not shown) of Mr. B Ob-2 who is one of the monitored persons Ob. The third sensor unit SUa-3 is disposed in a living room RM-3 (not shown) of Mr. C's Ob-3, who is one of the monitored persons Ob, and the fourth sensor unit SUa-4 is It is arrange | positioned by the room RM-4 (not shown) of Mr. D Ob-4 who is one of the to-be-monitored person Ob.

  In the present embodiment, the camera 12 corresponds to an example of an image acquisition unit, and the head detection processing unit 32 corresponds to an example of a head detection device. That is, the sensor device SUa includes a head detection device.

  The management server device SV has a communication function of communicating with the other devices SUa, SP, TA via the network NW, and receives from the sensor device SUa the detection result regarding the monitored person Ob and the image of the monitored person Ob. It is a device that manages information (monitoring information) related to monitoring of the monitored person Ob. When the management server device SV receives the detection result on the monitored person Ob and the image of the monitored person Ob from the sensor device SUa, the management server device stores (records) the monitoring information related to the monitoring of the monitored person Ob and A communication signal (monitoring information communication signal) containing the monitoring information relating to the monitoring of the supervisor Ob is transmitted to the fixed terminal apparatus SP and the portable terminal apparatus TA. Also, the management server SV provides the client with data in response to the request of the client (in the present embodiment, the fixed terminal SP and the portable terminal TA, etc.). Such a management server SV can be configured, for example, by a computer with a communication function.

  The fixed terminal device SP includes a communication function of communicating with other devices SUa, SV, TA via the network NW, a display function of displaying predetermined information, and an input function of inputting predetermined instructions and data. The user interface of the person-to-be-monitored monitoring system MS by inputting predetermined instructions and data given to the management server device SV and the portable terminal device TA, and displaying detection results and images obtained by the sensor device SUa It is a device that functions as (UI). Such fixed terminal apparatus SP can be configured, for example, by a computer with a communication function.

  The mobile terminal device TA has a communication function of communicating with other devices SV, SP, and SUa via the network NW, a display function of displaying predetermined information, an input function of inputting predetermined instructions and data, and a voice call. It has a call function to be performed, etc. and inputs predetermined instructions and data given to the management server SV and the sensor SUa, or displays the detection result and image obtained by the sensor SUa by notification from the management server SV It is an apparatus which receives and displays the said monitoring information regarding monitoring with respect to the to-be-monitored person Ob by doing etc. Such a portable terminal device TA can be configured by, for example, a portable communication terminal device such as a so-called tablet computer, a smartphone, or a mobile phone.

  Next, the operation of this embodiment will be described. In the monitored person monitoring system MSa having such a configuration, each of the devices SUa, SV, SP, and TA performs initialization of necessary parts when the power is turned on, and starts its operation. Further, in the sensor device SUa, the control processing unit 3 includes the control unit 31, the head detection processing unit 32, the action detection processing unit 33, the notification processing unit 34, the streaming processing unit 35, and the nurse by executing the control processing program. The call processing unit 36 is functionally configured, and in the head detection processing unit 32, the moving body extracting unit 321, the circular area extracting unit 322, the low reflectance area extracting unit 323, and the head selecting unit 324 are functionally configured. Be done.

  The monitored person monitoring system MSa having the above-described configuration monitors each of the monitored persons Ob generally by the following operation. The sensor unit SUa samples the output of the Doppler sensor 11 of the sensor unit 1 at a predetermined sampling cycle and acquires an image by the camera 12. The behavior detection processing unit 33 outputs the sampled and acquired Doppler sensor 11 and the camera A predetermined action (state, condition) in the monitored person Ob is determined based on the 12 images, and as a result of this determination, the predetermined action (for example, wake up in the present embodiment, When it is determined that the bed is falling, falling, a micro-motion abnormality, etc., determination result information indicating the determination result determined as the state of the monitored person Ob by the notification processing unit 34 or the image data of the still image of the monitored person Ob Communication signal (monitoring information communication signal) containing monitoring information such as the above is transmitted to the management server apparatus SV via the network NW. In addition, the operation (head detection operation) of detecting the head for detecting getting up, leaving, and falling will be described in detail later.

  When the management server SV receives the monitoring information communication signal from the sensor unit SUa via the network NW, the management server SV stores the monitoring information such as the determination result information and the still image data stored in the monitoring information communication signal. Store (Record) Then, the management server device SV transmits a monitoring information communication signal containing monitoring information such as the determination result information and the image data of the still image to the terminal device (the fixed terminal device SP and the portable terminal device TA in the present embodiment). . As a result, the status (condition) of the person to be monitored Ob is notified to the monitor such as a nurse or a care worker via the terminal device SP or TA.

  When the fixed terminal apparatus SP and the portable terminal apparatus TA receive the monitoring information communication signal from the management server apparatus SV via the network NW, they display the monitoring information contained in the monitoring information communication signal. According to such an operation, the monitored person monitoring system MS detects each monitored person Ob mainly by each sensor device SUa, the management server device SV, the fixed terminal device SP and the portable terminal device TA and detects each monitored person. Ob is monitored.

  Next, the head detection operation in the monitored person monitoring system MSa will be described. FIG. 4 is a flowchart showing the operation of the first aspect of the sensor device in the monitored person monitoring system of the first embodiment. FIG. 5 is a diagram for explaining an example of an image acquired by the sensor device in the monitored person monitoring system of the first embodiment. FIG. 6 is a diagram showing a pattern of a futon. FIG. 7 is a flowchart showing the operation of the second aspect of the sensor device in the monitored person monitoring system of the first embodiment. The sensor device SUa in this embodiment detects the head (head region) by the operation of the first mode shown in FIG. 4, for example, and also detects the head (head of the second device shown in FIG. 7). Area) is detected. Hereinafter, the operation of the first aspect will be described first, and then the operation of the second aspect will be described.

  First, the operation of the first aspect will be described. The sensor device SUa samples the output of the Doppler sensor 11 of the sensor unit 1 at each sampling timing, and operates as follows every time an image is acquired by the camera 12, thereby the head of the monitored person Ob ( The head region is detected, and a predetermined action in the monitored person Ob is detected.

  In FIG. 4, first, the control processing unit 3 acquires a Doppler signal from the Doppler sensor 11 of the sensor unit 1 and acquires an image (an image for one frame) from the camera 12 of the sensor unit 1 (S11). For example, an image shown in FIG. 5 is acquired by the camera 12. In the image shown in FIG. 5 of this example, not only the monitored person Ob consisting of the trunk BD and the substantially circular head HD but also the bedding BT, the substantially circular pillow PI on the bedding BT, and the bedding BT are hung The duvet BD, the substantially circular trash can WB and the substantially circular basin BA are shown. The futon BD has a circular pattern PT, and the circular pattern PT is shown in the image. In the case where the head HD is detected only by the process of extracting the substantially circular area from the image shown in FIG. 5, not only the head HD is detected, but also the pillow PI, the pattern PT, and the trash can WB And there is a possibility that basin BA may be misdetected as head HD. The pattern PT of the duvet BD may be erroneously detected as the head HD as well as the circular shape shown in FIG. 5, for example, the shape shown in FIG. In FIG. 6, an octagon (representative example of a polygon), a substantially circular shape surrounded by a V-shaped jagged edge, a substantially circular shape surrounded by eight triangles in a radial shape (shape simulating the sun), Three substantially circular shapes (three shapes imitating a flower pattern), a triangle and a quadrangle that are radially surrounded by a plurality of petals are illustrated. The shape of a triangle or a quadrilateral is distorted with the distortion of the duvet BD, so there is a possibility that it may be erroneously detected as a substantially circular shape. In order to reduce such false detections, in the present embodiment, the image is processed as follows. The reflectance to infrared light in the pillow PI, the pattern PT, the trash can WB and the basin BA is generally higher (larger) than the reflectance to infrared light on the head, depending on the material and the like.

  Returning to FIG. 4, next, the control processing unit 3 causes the moving object extraction unit 321 of the head detection processing unit 32 to use, for example, a method such as a background difference method or an interframe difference method from the image acquired by the camera 12. By using it, a moving body area extraction process is performed to extract a moving body area as a human body area of the monitored person Ob, and it is determined whether or not the moving body area is extracted (S12). As a result of this determination, when the moving body area is not extracted (No), the control processing unit 3 ends the processing at this timing, and when the moving body area is extracted (Yes) The control processing unit 3 notifies the moving body region extracted by the moving body extraction unit 321 from the moving body extraction unit 321 to the circular area extraction unit 322, and executes the next process S13.

  In this process S13, the control processing unit 3 uses, for example, a method such as Hough transformation or pattern matching from the moving object region extracted by the moving object extracting unit 321 by the circular region extracting unit 322 of the head detection processing unit 32. Then, a circular area extraction process of extracting a substantially circular area (circular area) is executed to determine whether the circular area is extracted. As a result of this determination, when the circular area is not extracted (No), the control processing unit 3 ends the processing at this timing, and when the circular area is extracted (Yes) The control processing unit 3 notifies the circular area extraction unit 322 of the circular area extracted by the circular area extraction unit 322 from the circular area extraction unit 322 to the low reflectance area extraction unit 323 and the head selection unit 324, and performs the next process Execute S14.

  In the process S14, the control processing unit 3 causes the head selection unit 324 of the head detection processing unit 32 to determine whether the number of circular regions extracted by the circular region extraction unit 322 is more than one. As a result of this determination, when the number of the circular regions is not plural, that is, when the number of the circular regions is one (No), the control processing unit 3 causes the head selection unit 324 to execute the process S13. The extracted circular area is selected as the head (head area), and the head (head area) selected by the head selection unit 324 is sent from the head detection processing unit 32 to the action detection processing unit 33 It notifies and performs the following processing S15. On the other hand, as a result of the determination, if the circular area is plural (Yes), the control processing unit 3 determines that the head (head area) is in the plural circular areas extracted in the process S13. It is assumed that an object other than is also extracted, and the next process S17 is executed.

  In the process S17, the control processing unit 3 causes the low reflectance area extracting unit 323 of the head detection processing unit 32 to generate an area equal to or less than the threshold from the circular area extracted by the circular area extracting unit 322 in the process S13. The low reflectance area extraction process of extracting the low reflectance area as the low reflectance area is performed, and the next process S18 is performed.

  In the process S18, the control processing unit 3 determines whether or not the number of low reflectance areas extracted by the low reflectance area extraction unit 323 is one by the head selection unit 324 of the head detection processing unit 32. If it is determined that the number of the low reflectance areas is one as a result of this determination (Yes), the control processing unit 3 causes the head selection unit 324 to display the low reflectance areas extracted in the process S17. The head detection processing unit 32 notifies the behavior detection processing unit 33 of the head (head region) selected by the head selection unit 324, and executes the next process S15. . On the other hand, as a result of the determination, when the number of the low reflectance areas is not one, that is, when the number of the low reflectance areas is plural (No), the control processing unit 3 extracts the plural extracted in the process S17. It is assumed that an object that is not the head (head region) is also extracted in the low reflectance region of the above, and the head selection unit 324 determines a predetermined threshold (for example, 50%, 30%, 10%, etc.) The threshold value is reset by subtracting a value (for example, 0.5%, 0.2%, etc.) (S19), and the process returns to the process S17. Therefore, the control processing unit 3 changes the predetermined threshold value from the default threshold value by the head selection unit 324, thereby performing the low reflectance area extraction process by the low reflectance area extraction unit 323 in the process S17. The processes S17 to S19 are repeated until there is one low reflectance area.

  For example, in the image shown in FIG. 5 described above, not only the head HD but also the above-described processing S12 and processing S13 by shifting the pillow PI and the duvet BD and moving the trash can WB and the basin BA. , Pillow PI, pattern PT, trash can WB and basin BA are also extracted as circular shaped areas. If the pattern PT is erroneously detected as a head (head area), the fall of the duvet BD may be erroneously detected as falling of the monitored person Ob. However, in the present embodiment, the processes S17 to S19 are performed on the plurality of circular regions through the execution of the process S14. Here, the pillow PI, the pattern PT, the trash can WB and the basin BA are made of a material (for example, fabric, metal, metal coating, plastic, rattan, etc.), and the infrared reflectance is red at the head Since the reflectance for external light is substantially higher (larger), the execution of the processing S17 to the processing S19 removes it from the low reflectance region, leaving only the head HD as the low reflectance region. By thus considering the reflectance to infrared light, the head HD can be detected more accurately.

  Referring back to FIG. 4, in the processing S15, the head (head region) is detected by the above-described processing S11 to S14 and S17 to S19. Based on the output of the unit 1, an action detection process for detecting a predetermined action set in advance in the monitored person Ob is executed, and it is determined whether or not the predetermined action is detected. More specifically, the action detection processing unit 33 selects the person to be monitored Ob from the height (size) of the head (head area) detected by the head detection processing unit 32 as one aspect of the action detection processing. Posture (for example, standing position, sitting position, lying position, etc.), the position of the detected head is detected, and based on these positions, the posture and the position of the person to be monitored Ob detected Determine another. The action detection processing unit 33 detects body movement (up and down movement of the chest) associated with the breathing operation of the person Ob to be monitored by the Doppler sensor 11 of the sensor unit 1 as another aspect of the action detection processing. Determine if there is any movement abnormality. As a result of the determination, if the predetermined action is not detected (No), the control processing unit 3 ends the processing at this timing, while the predetermined action (in this embodiment, getting up, If at least one of bed leaving, falling, and body movement abnormality is detected (Yes), the control processing unit 3 notifies the notification processing unit 34 of the predetermined action detected in the processing S15, The next process S16 is performed.

  In this process S16, when the predetermined action in the monitored person Ob is notified from the action detection processing unit 33, the notification processing unit 34 notifies that to the management server SV by the above-mentioned monitoring information communication signal, The processing at this timing ends.

  Next, the operation of the second aspect will be described. The sensor device SUa samples the output of the Doppler sensor 11 of the sensor unit 1 at each sampling timing, and operates as follows every time an image is acquired by the camera 12, thereby the head of the monitored person Ob ( The head region is detected, and a predetermined action in the monitored person Ob is detected.

  In FIG. 7, first, the control processing unit 3 acquires a Doppler signal from the Doppler sensor 11 of the sensor unit 1 in the same manner as the process S11 described above, and an image (image for one frame) from the camera 12 of the sensor unit 1. Acquire (S21).

  Next, the control processing unit 3 executes the moving body region extraction processing by the moving body extraction unit 321 of the head detection processing unit 32 and determines whether the moving body region is extracted, as in the above-described processing S12. To do (S22). As a result of this determination, when the moving body area is not extracted (No), the control processing unit 3 ends the processing at this timing, and when the moving body area is extracted (Yes) The control processing unit 3 notifies the low-reflectance area extraction unit 323 of the moving body area extracted by the moving body extraction unit 321 from the moving body extraction unit 321, and executes the next process S23.

  In the processing S23, the control processing unit 3 causes the low reflectance area extraction unit 323 of the head detection processing unit 32 to reduce the reflection of the area below the threshold from the moving body area extracted by the moving body extraction unit 321 in the process S22. Low-reflectance area extraction processing for extracting the low-reflectance area extracted as the low-reflectance area extraction unit 323 is notified from the low-reflectance area extraction unit 323 to the circular area extraction unit 322. Process S24 of FIG.

  In the process S24, the control processing unit 3 performs, for example, Hough transformation, pattern matching, and the like from the low reflectance area extracted by the low reflectance area extraction unit 323 by the circular area extraction unit 322 of the head detection processing unit 32. A circular area extraction process of extracting a substantially circular area (circular area) is executed by using the method of (4), and it is determined whether the circular area is extracted. As a result of this determination, when the circular area is not extracted (No), the control processing unit 3 ends the processing at this timing, and when the circular area is extracted (Yes) The control processing unit 3 notifies the circular area extracted by the circular area extraction unit 322 from the circular area extraction unit 322 to the head selection unit 324, and executes the next process S25.

  In the process S25, the control processing unit 3 causes the head selection unit 324 of the head detection processing unit 32 to determine whether or not there is one circular area extracted by the circular area extraction unit 322. If it is determined that the number of circular regions is one as a result of this determination (Yes), the control processing unit 3 causes the head selecting unit 324 to select the circular regions extracted in the process S24 as the head (head The head detection processing unit 32 notifies the behavior detection processing unit 33 of the head (head region) selected by the head selection unit 324, and the next process S26 is executed. On the other hand, as a result of the determination, when the number of circular regions is not one, that is, when the number of circular regions is plural (No), the control processing unit 3 extracts the plurality of the extracted in the process S24. It is assumed that an object that is not the head (head region) is also extracted in the circular region, and the head selection unit 324 extracts a predetermined value (for example, 50%, 30%, 10%, etc.) The threshold is reset by subtracting 0.5%, 0.2%, etc.) (S28), and the process returns to step S23. Therefore, the control processing unit 3 changes the predetermined threshold value from the default threshold value by the head selection unit 324 so that the low reflectance area extraction process by the low reflectance area extraction unit 323 in the process S23 is performed. The processes S23 to S25 and S28 are repeated until the area is a single circular area.

  On the other hand, in the processing S26, since the head (head region) is detected by the above-described processing S21 to S25 and S28, the control processing unit 3 executes the action detection processing in the same manner as the processing S15 described above. And determine whether the predetermined action has been detected. As a result of the determination, if the predetermined action is not detected (No), the control processing unit 3 ends the processing at this timing, while the predetermined action (in this embodiment, getting up, If at least one of bed leaving, falling, and body movement abnormality is detected (Yes), the control processing unit 3 notifies the notification processing unit 34 of the predetermined action detected in the processing S26, The next process S27 is executed.

  In the process S27, when the action detection processing unit 33 is notified of the predetermined action in the monitored person Ob, the notification processing unit 34 notifies that by the monitoring information communication signal as in the above-described process S16. The management server SV is notified, and the processing at this timing is ended.

  As described above, the monitored person monitoring system MSa which is an example of the monitored person monitoring device, the head detection device, and the sensor device SUa of the example using the head detection method are not only substantially circular but also infrared. Since the head HD is detected in consideration of the reflectance with respect to light, the head HD of the monitored person Ob can be detected more accurately based on the image captured from the upper side of the monitored person Ob. As a result, the to-be-monitored person monitoring system MSa and the sensor device SUa can detect a predetermined action in the to-be-monitored person Ob more accurately.

  Further, in the operation of the above-described first aspect, the monitored person monitoring system MSa and the sensor device SUa narrow down the circular area extracted by the circular area extraction unit 322 as a head area candidate, and use the narrowed down candidates Since the head HD is detected in consideration of the reflectance to infrared light, it is possible to simply erroneously detect one having a low reflectance as compared with the case of extracting a low reflectance region from the entire image acquired by the camera 12 It can be reduced.

  Further, in the operation of the second aspect described above, the monitored person monitoring system MSa and the sensor device SUa narrow down the low reflectance area extracted by the low reflectance area extraction unit 323 as a head area candidate, and narrow down this Since the head HD is detected in consideration of the substantially circular shape from the candidates, the amount of information processing can be reduced compared to the case where a substantially circular circular region is extracted from the entire image acquired by the camera 12.

  In the above-described embodiment, in order to further reduce false detection of something that approximates a head that is not a head, a predetermined marker may be added to something that approximates a head that is not a head. In this case, the head detection processing unit 32 of the sensor device SUa, which is an example of a head detection device, is obtained by the camera 12 of the sensor unit 1, which is an example of an image acquisition unit, as shown by a dashed line in FIG. The head selecting unit 324 is functionally further provided with a marker detecting unit 325 for detecting a predetermined marker based on the image, and the head selecting unit 324 includes a circular region extracted by the circular region extracting unit 322 and a low reflectance region extracting unit 323. An area of the head is selected based on the extracted low reflectance area and the marker detected by the marker detection unit 325 to detect the head. Such a sensor device SUa can also detect the head of the person Ob who is to be monitored with higher accuracy because it also considers markers.

  More specifically, the head selection unit 324 detects markers detected by the marker detection unit 325 from the circular area extracted by the circular area extraction unit 322 and the low reflectance area extracted by the low reflectance area extraction unit 323. An area of the head is selected based on the remaining circular area excluding the area having the area and the low reflectance area to detect the head. In this case, for example, in the processing S12 described above, the remaining processing areas after removal of the area having the marker detected by the marker detection section 325 from the moving body area extracted by the moving body extraction section 321 S13 to S19 are executed. Also, for example, in the above-described processing S13, the remaining circular regions excluding the region having the marker detected by the marker detection unit 325 from the circular region extracted by the circular region extraction unit 322 are described above Each process S14 to S19 is executed. Also, for example, with respect to the low reflectance area remaining after excluding the area having the marker detected by the marker detection unit 325 from the low reflectance area extracted by the low reflectance area extraction unit 323 in the above-described processing S17. The subsequent processes S15 to S19 described above are performed. Further, for example, in the process S22 described above, the remaining processes in the moving body area excluding the area having the marker detected by the marker detection unit 325 from the moving body area extracted by the moving body extraction unit 321 S28 is executed. Also, for example, with respect to the low reflectance area remaining after excluding the area having the marker detected by the marker detection unit 325 from the low reflectance area extracted by the low reflectance area extraction unit 323 in the above-described processing S23. The subsequent processes S23 to S28 described above are performed. Further, for example, in the above-described processing S24, the remaining circular regions excluding the region having the marker detected by the marker detection unit 325 from the circular region extracted by the circular region extraction unit 322 are described above The respective processes S23 to S28 are executed. By providing a marker to a head approximating a head that is not the head, such a sensor device SUa removes the circular area having the marker and the low reflectance area, thereby forming a circular area extraction unit. From the circular area extracted in 322 and the low reflectance area extracted by the low reflectance area extraction unit 323, the non-head part can be removed, and the head of the monitored person Ob can be detected more accurately.

  Such a marker is preferably a marker having a specific shape. For example, a marker having, for example, a rectangular shape or a circular arc shape is attached to the edge of the bedding BT such as the edge of the trash can WB or the bed. Such a sensor device SUa can reduce false detection in the marker detection unit 325 by using a marker having a specific shape. Also preferably, the marker is a marker distinguishable only by infrared light. Preferably, the marker distinguishable only by the infrared light is a material that reflects infrared light with a reflectance (for example, a reflectance of 50% or more) higher than the reflectance of the head to the infrared light, the material In the seal (eg, metal seal, wood seal, etc.) formed by the above, a pigment (infrared reflection pigment) that reflects infrared light with a reflectance (eg, 50% or higher reflectance) higher than the reflectance of the head to infrared light And a combination of one or more of the predetermined shapes formed by the difference in the density of the fabric weave. For example, the duvet BD or pillow PI is made of the above-mentioned material that reflects infrared light with a high reflectance, whereby the duvet BD or pillow PI is provided with a marker (the duvet BD itself or the pillow PI itself is a marker ). Also, for example, since the reflectance to infrared light differs between natural materials such as cotton and hemp and chemical fiber materials such as polyester, chemical fibers can be used for the duvet BD and pillow PI made of natural materials (or chemical fiber materials). By sewing a thread of material (or natural material), a marker is given to the duvet BD and the pillow PI. Also, for example, since the density of the woven fabric differs in reflectance to infrared light, the fabric of the predetermined shape of the woven fabric (or the duvet BD or pillow PI made of the woven rough fabric (or the dense fabric of the woven fabric) A marker is given to the duvet BD and the pillow PI by sewing in a woven rough fabric). Such a sensor apparatus SUa can reduce the erroneous detection in the marker detection part 325 by using the marker which can be identified only with infrared light. Since the marker can not be seen, the sensor device SU does not give a sense of discomfort due to the marker without causing the monitored person Ob to recognize the presence of the marker when used in the monitored person monitoring system MSa.

Next, another embodiment will be described.
Second Embodiment
FIG. 8 is a diagram showing a configuration of a sensor device in the monitored person monitoring system of the second embodiment.

  The monitored person monitoring system MSb in the second embodiment of the example in which the monitored person monitoring device is realized as a system is monitoring (should be watched) monitoring in the same manner as the monitored person monitoring system MSa in the first embodiment. The monitoring target person Ob is detected by detecting a predetermined behavior set in advance in the monitoring target person (watching target person) Ob (Ob-1 to Ob-4) which is a target (watching target), For example, one or more sensor devices SUb (SUb-1 to SUb-4), management server device SV, fixed terminal device SP, and one or more mobile terminal devices TA (TA-1 and TA-2) Equipped with That is, the monitored person monitoring system MSb in the second embodiment is the same as the monitored person monitoring system MSa in the first embodiment except that the sensor device SUa is provided instead of the sensor device SUa. Therefore, the descriptions of the management server device SV, the fixed terminal device SP, and the portable terminal device TA will be omitted.

  Similar to the sensor device SUa, the sensor device SUb has a communication function of communicating with other devices SV, SP, and TA via the network NW, detects the monitored person Ob, and sends the detection result to the management server device SV. It is an apparatus to transmit. For example, as shown in FIG. 8, such a sensor device SUb includes a sensor unit 1, a sound input / output unit 2, a control processing unit 6, a communication interface unit (communication IF unit) 4, and a storage unit 7. Equipped with That is, the sensor device SUb is the same as the sensor device SUa except that it includes a control processing unit 6 instead of the control processing unit 3 and a storage unit 7 instead of the storage unit 5. Therefore, the description of the sensor unit 1, the sound input / output unit 2 and the communication IF unit 4 is omitted.

  The storage unit 7 is a circuit which is connected to the control processing unit 6 and stores various predetermined programs and various predetermined data according to the control of the control processing unit 6. The various predetermined programs include, for example, a control processing program such as a monitoring processing program that executes information processing related to monitoring of the monitored person Ob. The monitoring processing program includes a head detection processing program for detecting the head of the person to be monitored Ob based on the output of the sensor unit 1 and a predetermined action on the person to be monitored Ob based on the output of the sensor unit 1. An action detection processing program, a notification processing program for notifying the predetermined action detected by the action detection program to the outside, a moving image captured by the camera 12 by streaming to the fixed terminal apparatus SP or the portable terminal apparatus TA that requested the moving image It includes a streaming processing program to be distributed, and a nurse call processing program for voice communication with the fixed terminal apparatus SP or the portable terminal apparatus TA by using the sound input / output unit 2 or the like. The head detection processing program includes a moving body extraction program for extracting a moving body area as a human body area of the monitored person Ob from an image acquired by the camera 12 of the sensor unit 1, and a moving body area extracted by the moving body extraction program. A circular area extraction program for extracting a substantially circular area as a circular area, a marker detection program for detecting a predetermined marker based on the moving body area extracted by the moving body extraction program, and the circular area extraction And a head selection program for selecting the head region based on the circular region extracted by the program and the marker extracted by the marker detection program. The action detection processing program, as one aspect, detects a predetermined action in the monitored person Ob based on the head of the monitored person Ob detected by the head detection processing program. The various predetermined data includes data necessary for executing each of these programs, data necessary for monitoring the monitored person Ob, and the like. The storage unit 7 includes, for example, a ROM, which is a nonvolatile storage element, and an EEPROM, which is a rewritable nonvolatile storage element. The storage unit 7 includes, for example, a RAM serving as a working memory of the so-called control processing unit 3 that stores data and the like generated during execution of the predetermined program.

  The control processing unit 6 controls the respective units of the sensor device SUb according to the functions of the respective units, and the sensor unit 1 detects predetermined amounts set in advance in the person to be monitored Ob, and the detected predetermined amount It is a circuit for detecting and notifying a predetermined action in the person to be monitored Ob based on various amounts. The control processing unit 6 includes, for example, a CPU and its peripheral circuits. The control processing unit 6 executes the control processing program to control the control unit 61, the head detection processing unit 62, the action detection processing unit 63, the notification processing unit 64, the streaming processing unit 65, and the nurse call processing unit 66. Functionally equipped. The control unit 61, the action detection processing unit 63, the notification processing unit 64, the streaming processing unit 65, and the nurse call processing unit 66 in the control processing unit 6 respectively have a control unit 31 in the control processing unit 3 of the first embodiment and an action. Since the detection processing unit 33, the notification processing unit 34, the streaming processing unit 35, and the nurse call processing unit 36 are the same as the detection processing unit 33, the description thereof will be omitted.

  The head detection processing unit 62 detects the head of the monitored person Ob based on the output of the sensor unit 1 as the head detection processing unit 32 does. More specifically, the head detection processing unit 62 detects the head of the monitored person Ob based on the image captured by the camera 12 of the sensor unit 1 and captured from above the monitored person Ob. To this end, the head detection processing unit 62 functionally includes a moving body extraction unit 621, a circular area extraction unit 632, a marker detection unit 623, and a head selection unit 624.

  Similar to the moving body extraction unit 321, the body movement extraction unit 621 uses a method such as a background difference method or an interframe difference method from an image acquired by the camera 12 of the sensor unit 1 as a human body region of the monitored person Ob. It is to extract a moving body area. The moving body extracting unit 621 notifies the extracted moving body region to the circular area extracting unit 622 in the operation of the first mode described later, and notifies the marker detecting unit 623 in the operation of the second mode described later.

  The circular area extraction unit 622 extracts a substantially circular area as a circular area based on the image acquired by the camera 12 of the sensor unit 1 as the circular area extraction unit 322 does. More specifically, in the operation of the first aspect, the circular area extraction unit 622 extracts a circular area from the moving body area extracted by the moving body extraction unit 621, for example, by the Hough transform method or the pattern matching method. The extracted circular area is notified to the marker detection unit 623. Further, in the operation of the second aspect, the circular area extraction unit 322 performs, for example, the Hough transform method or the pattern matching from the remaining body movement area excluding the area of the marker detected by the marker detection unit 623 as described later. A circular area is extracted by the method or the like, and the extracted circular area is notified to the head selecting unit 624.

  The marker detection unit 623 detects a predetermined marker based on the image acquired by the camera 12 of the sensor unit 1. The above-mentioned predetermined marker is given to what approximates a head which is not a head in order to reduce false detection of what approximates a head which is not a head, as described above, and preferably, a specific shape The marker is preferably a marker that is distinguishable only by infrared light. More specifically, in the operation of the first aspect, the marker detection unit 623 detects the predetermined marker from the circular area extracted by the circular area extraction unit 622, and the circular area extraction unit 622 extracts the predetermined marker. The remaining circular area excluding the circular area having the detected marker is notified from the circular area to the head selection unit 624. Further, in the operation of the second aspect, the marker detection unit 623 detects the predetermined marker from the moving subject region extracted by the moving subject extraction unit 621 and has the detected marker from the moving subject region extracted by the moving subject extraction unit 621. The remaining body movement area excluding the body movement area is notified to the circular area extraction unit 622.

  The head selection unit 624 selects the area of the head based on the circular area extracted by the circular area extraction unit 622 and the predetermined marker extracted by the marker detection unit 623, and detects the head. It is. More specifically, in the operation of the first aspect, the head selection unit 624 selects the region of the head based on the marker detected by the marker detection unit 623, detects the head, and detects the head. The part (head area) is notified to the action detection processing part 63. Further, in the operation of the second aspect, the head selection unit 624 selects the area of the head based on the circular area extracted by the circular area extraction unit 622, detects the head, and detects the head. The part (head area) is notified to the action detection processing part 63.

  In the monitored person monitoring system MSb having such a configuration, each of the devices SUb, SV, SP, and TA performs initialization of necessary parts when the power is turned on, and starts its operation. Further, in the sensor device SUb, the control processing unit 6 includes the control unit 61, the head detection processing unit 62, the action detection processing unit 63, the notification processing unit 64, the streaming processing unit 65, and the nurse by executing the control processing program. The call processing unit 66 is functionally configured, and in the head detection processing unit 62, the moving body extracting unit 621, the circular area extracting unit 622, the low reflectance area extracting unit 623 and the head selecting unit 624 are functionally configured Be done. The monitored person monitoring system MSb having the above configuration operates in the same manner as the monitored person monitoring system MSa described above except that the sensor device SUb operates as described later, and thus the description thereof will be omitted.

  Next, the head detection operation in the monitored person monitoring system MSb will be described. FIG. 9 is a flowchart showing the operation of the first aspect of the sensor device in the monitored person monitoring system of the second embodiment. FIG. 10 is a flowchart showing the operation of the second aspect of the sensor device in the monitored person monitoring system of the second embodiment. The sensor device SUb in the present embodiment detects the head (head area) by the operation of the first mode shown in FIG. 9, for example, and also detects the head (head of the second mode shown in FIG. 10). Area) is detected. Hereinafter, the operation of the first aspect will be described first, and then the operation of the second aspect will be described.

  First, the operation of the first aspect will be described. The sensor device SUb samples the output of the Doppler sensor 11 of the sensor unit 1 at each sampling timing, and operates as follows every time an image is acquired by the camera 12, thereby the head of the monitored person Ob ( The head region is detected, and a predetermined action in the monitored person Ob is detected.

  In FIG. 9, first, the control processing unit 6 acquires a Doppler signal from the Doppler sensor 11 of the sensor unit 1 in the same manner as the above-described processing S11, and an image (image for one frame) from the camera 12 of the sensor unit 1 Acquire (S31).

  Next, the control processing unit 6 executes the moving body region extraction processing by the moving body extraction unit 621 of the head detection processing unit 62 and determines whether the moving body region is extracted, as in the above-described process S12. (S32). As a result of this determination, when the moving body area is not extracted (No), the control processing unit 6 ends the processing at this timing, and when the moving body area is extracted (Yes) The control processing unit 6 notifies the moving body region extracted by the moving body extraction unit 621 from the moving body extraction unit 621 to the circular area extraction unit 622, and executes the next process S33.

  In this process S33, the control processing unit 6 uses the method such as Hough transformation or pattern matching from the moving object region extracted by the moving object extracting unit 321 by the circular region extracting unit 622 of the head detection processing unit 62. Then, a circular area extraction process of extracting a substantially circular area (circular area) is executed to determine whether the circular area is extracted. As a result of this determination, when the circular area is not extracted (No), the control processing unit 6 ends the processing at this timing, and when the circular area is extracted (Yes) The control processing unit 6 notifies the circular area extraction unit 622 of the circular area extracted by the circular area extraction unit 622 to the marker detection unit 623 and the head selection unit 624, and executes the next process S34. Do.

  In the process S34, the control processing unit 6 determines, by the head selection unit 624 of the head detection processing unit 62, whether or not the number of circular regions extracted by the circular region extraction unit 622 is more than one. As a result of this determination, when the number of the circular regions is not plural, that is, when the number of the circular regions is one (No), the control processing unit 6 causes the head selection unit 624 to execute the process S33. The extracted circular area is selected as the head (head area), and the head (head area) selected by the head selection unit 624 is sent from the head detection processing unit 62 to the action detection processing unit 63. It notifies and performs the next process S35. On the other hand, as a result of the determination, if the circular area is plural (Yes), the control processing unit 6 determines that the head (head area) is included in the plural circular areas extracted in the process S33. It is assumed that an object that is not is extracted, and the next process S37 is executed.

  In step S37, the control processing unit 6 causes the marker detection unit 623 of the head detection processing unit 62 to detect a predetermined marker from the circular area extracted by the circular area extraction unit 622, and the circular area extraction unit The remaining circular area excluding the circular area having the detected marker is notified from the circular area extracted at 622 to the head selecting unit 624, and the remaining part removed by the marker detecting unit 623 by the head selecting unit 624 The circular area is selected as the area of the head to detect the head, the detected head (head area) is notified to the action detection processing unit 63, and the next process S35 is executed. The control processing unit 6 may notify the behavior detection processing unit 63 from the marker detection unit 623 as the remaining circular area removed by the marker detection unit 623 as the area of the head.

  In the process S35, since the head (head region) is detected by the processes S31 to S34 and S37 described above, the control processing unit 6 causes the behavior detection processing unit 63 to perform the process in the same manner as the process S15 described above. Based on the output of the sensor unit 1, an action detection process for detecting a predetermined action set in advance in the monitored person Ob is executed, and it is determined whether or not the predetermined action has been detected. As a result of the determination, if the predetermined action is not detected (No), the control processing unit 6 ends the processing at this timing, while the predetermined action (in this embodiment, getting up, If at least one of bed leaving, falling, and abnormal body movement is detected (Yes), the control processing unit 6 notifies the notification processing unit 64 of the predetermined action detected in the processing S35, The next process S36 is executed.

  In this process S36, when the predetermined action in the monitored person Ob is notified from the action detection processing unit 63, the notification processing unit 64 notifies that to the management server SV by the above-mentioned monitoring information communication signal, The processing at this timing ends.

  Next, the operation of the second aspect will be described. The sensor device SUb samples the output of the Doppler sensor 11 of the sensor unit 1 at each sampling timing, and operates as follows every time an image is acquired by the camera 12, thereby the head of the monitored person Ob ( The head region is detected, and a predetermined action in the monitored person Ob is detected.

  In FIG. 10, first, the control processing unit 6 acquires a Doppler signal from the Doppler sensor 11 of the sensor unit 1 in the same manner as the above-described processing S11, and an image (image for one frame) from the camera 12 of the sensor unit 1. Acquire (S41).

  Next, the control processing unit 6 executes the moving body region extraction processing by the moving body extraction unit 621 of the head detection processing unit 62 and determines whether the moving body region is extracted, as in the above-described process S12. (S42). As a result of this determination, when the moving body area is not extracted (No), the control processing unit 6 ends the processing at this timing, and when the moving body area is extracted (Yes) The control processing unit 6 notifies the moving body region extracted by the moving body extraction unit 621 from the moving body extraction unit 621 to the marker detection unit 623, and executes the next process S43.

  In the processing S43, the control processing unit 6 causes the marker detection unit 623 of the head detection processing unit 62 to detect a predetermined marker from the body movement area extracted by the body movement extraction unit 621, and the body movement extraction unit 621 The remaining body movement area excluding the body movement area having the detected marker is notified from the extracted body movement area to the circular area extraction unit 622, and the next process S44 is performed.

  In the processing S44, the control processing unit 6 causes the circular area extraction unit 622 of the head detection processing unit 62 to, for example, carry out the Hough from the remaining body movement area excluding the body movement area having the marker by the marker detection unit 623. A circular area extraction process for extracting a substantially circular area (circular area) is executed by using a method such as conversion or pattern matching, and it is determined whether the circular area is extracted. As a result of this determination, when the circular area is not extracted (No), the control processing unit 6 ends the processing at this timing, and when the circular area is extracted (Yes) The control processing unit 6 notifies the circular area extraction unit 622 of the circular area extracted by the circular area extraction unit 622 from the circular area extraction unit 622 to the head selection unit 624, and the head selection unit 624 transmits the circular area extraction unit The circular area extracted by 622 is selected as an area of the head, the head is detected, the detected head (head area) is notified to the action detection processing unit 63, and the next process S45 is executed. Do. The control processing unit 6 may notify the behavior detection processing unit 63 from the circular region extraction unit 622 as the region of the head of the circular region extracted by the circular region extraction unit 622.

  In this process S45, since the head (head area) is detected by the processes S41 to S44 described above, the control processing unit 6 detects the sensor unit by the action detection processing unit 63 in the same manner as the process S15 described above. Based on the output of 1, an action detection process for detecting a predetermined action set in advance in the monitored person Ob is executed, and it is determined whether or not the predetermined action has been detected. As a result of the determination, if the predetermined action is not detected (No), the control processing unit 6 ends the processing at this timing, while the predetermined action (in this embodiment, getting up, If at least one of bed leaving, falling, and body movement abnormality) is detected (Yes), the control processing unit 6 notifies the notification processing unit 64 of the predetermined action detected in the processing S45, The next process S46 is performed.

  In the process S46, when the action detection processing unit 63 is notified of the predetermined action in the monitored person Ob, the notification processing unit 64 notifies the management server SV of the fact by the monitoring information communication signal. The processing at this timing ends.

  As described above, the monitored person monitoring system MSb, which is an example of the monitored person monitoring device, the head detection device, and the sensor device SUb of the example using the head detection method are not only substantially circular but also markers Since the head HD is detected in consideration, the head HD of the person to be monitored Ob can be detected more accurately based on the image obtained by capturing the person to be monitored Ob from above. As a result, the to-be-monitored person monitoring system MSb and the sensor device SUb can detect predetermined actions in the to-be-monitored person Ob more accurately.

  Further, in the operation of the first aspect described above, the monitored person monitoring system MSb and the sensor device SUb narrow down the circular shaped area extracted by the circular shaped area extracting unit 622 as the head area candidate, and use the narrowed down candidates By excluding the circular area having the marker, the circular area extracted by the circular area extraction unit 622 can exclude the non-head from the circular area, and the head HD of the person to be monitored Ob can be more accurately detected. It can be detected.

  Further, in the operation of the second aspect described above, the monitored person monitoring system MSb and the sensor device SUb remove the body motion area by removing the body motion area having the marker from the body motion area extracted by the body motion extracting unit 621. Since the head HD is detected in consideration of the substantially circular shape from the narrowed body movement region, the amount of information processing is compared to the case of extracting the substantially circular circular region from the entire image acquired by the camera 12 Can be reduced.

  Further, in the above-described embodiment, in the case where the circular area is detected within the area of the bedding BT, the circular area is the process S15, the process S26, the process S26, in view of the safety on the bedding BT. You may exclude from the object of the said action detection process in each of S35 and process S45.

  Moreover, in these above-mentioned embodiment, although sensor apparatus SUa and SUb were provided with the said action detection part and the said notification part, the said action detection part and the said notification part may be equipped in the management server apparatus SV, and The fixed terminal apparatus SP may be provided, or the mobile terminal apparatus TA may be provided. In such a case, a communication interface or the like that receives and acquires a communication signal containing an image and a Doppler signal captured from above the monitored person Ob from the sensor devices SUa and SUb via the network NW is an example of an image acquisition unit And one example of part of the distance measuring unit.

  While the present invention has been properly and sufficiently described above through the embodiments with reference to the drawings in order to express the present invention, those skilled in the art can easily change and / or improve the above embodiments. It should be recognized that it is possible. Therefore, unless a change or improvement implemented by a person skilled in the art is at a level that deviates from the scope of the claims set forth in the claims, the change or the improvement is the scope of the rights of the claim It is interpreted as being included in

Ob (Ob-1 to Ob-4) Monitored person NS (NS-1, NS-2) Observer (user)
MSa, MSb Monitored person monitoring system SUa, SUb Sensor device SV Management server device SP Fixed terminal device TA Portable terminal device 1 Sensor unit 3, 6 Control processing unit 4 Communication interface unit (communication IF unit)
5, 7 storage unit 11 Doppler sensor 12 camera (an example of an image acquisition unit)
31 and 61 control unit 32 and 62 head detection processing unit 33 and 63 action detection processing unit 34 and 64 notification processing unit 321 and 621 moving body extraction unit 322 and 622 circular area extraction unit 323 low reflectance area extraction unit 324 and 624 Head selection unit 325, 623 Marker detection unit

Claims (6)

An image acquisition unit for acquiring an image obtained by capturing a person to be monitored from above the person to be monitored;
A circular area extraction unit that extracts a substantially circular area as a circular area based on the image acquired by the image acquisition unit;
A low reflectance area extraction unit that extracts a region having a predetermined reflectance or less with respect to infrared light as a low reflectance region based on the image acquired by the image acquisition unit;
A marker detection unit configured to detect a predetermined marker based on the image acquired by the image acquisition unit;
The head region is selected based on the circular area extracted by the circular area extraction unit, the low reflectance area extracted by the low reflectance area extraction unit, and the marker detected by the marker detection unit. And a head selection unit for detecting
The head selecting unit is a circular area extracted by the circular area extracting unit, and a low reflectance area extracted by the low reflectance area extracting unit, the remaining area excluding the area having the marker detected by the marker detecting unit. Selecting the area of the head based on the circular area and the low reflectance area, and detecting the head;
The head detection device, wherein the marker is distinguishable only by the infrared light, and the reflectance to the infrared light exceeds the predetermined reflectance. The head according to claim 1, wherein the low reflectance area extraction unit extracts an area having a predetermined reflectance or less from the circular area extracted by the circular area extraction unit as a low reflectance area. Part detection device. The head detection according to claim 1, wherein the circular area extraction unit extracts a substantially circular area as a circular area from the low reflectance area extracted by the low reflectance area extraction section. apparatus. The head detection device according to any one of claims 1 to 3, wherein the marker is a marker having a specific shape. An image acquisition step of acquiring an image obtained by photographing a person to be monitored from above the person to be monitored;
A circular area extraction step of extracting a substantially circular area as a circular area based on the image acquired in the image acquisition step;
A low reflectance area extraction process of extracting an area having a predetermined reflectance or less with respect to infrared light as a low reflectance area based on the image acquired in the image acquisition process;
A head selecting step of selecting a head region based on the circular shaped region extracted in the circular shaped region extracting step and the low reflectance region extracted in the low reflectance region extracting step;
And a marker detection step of detecting a predetermined marker based on the image acquired in the image acquisition step.
The head selecting step is the remaining portion excluding the region having the marker detected in the marker detecting step from the circular shaped region extracted in the circular shaped region extracting step and the low reflectance region extracted in the low reflectance region extracting step Selecting the area of the head based on the circular area and the low reflectance area, and detecting the head;
The head detection method, wherein the marker is distinguishable only by the infrared light, and the reflectance to the infrared light exceeds the predetermined reflectance. A head detection unit that detects a head of an image of a person to be monitored from an image captured from above the person to be monitored;
An activity detection unit that detects a predetermined activity of the monitored person based on the head of the monitored person detected by the head detection unit;
A notification unit for notifying the outside of the predetermined action detected by the action detection unit;
The said head detection part contains the head detection apparatus of any one of Claims 1-4, The to-be-monitored person monitoring apparatus characterized by these.

Images