JPH0894400A - Image processor and applied apparatus using the same - Google Patents

Abstract

PURPOSE: To accurately divide a plurality of persons to individual areas when the detected one human body area is the superposition of the plurality of the persons by processing an infrared image obtained from a two-dimensional infrared sensor, and accurately detecting the area of the body in the detecting area. CONSTITUTION: The image processor comprises a sensor 1 for measuring the temperatures of a person and an environment in a detecting area to output a thermal image signal, a background temperature calculator 2 for calculating the background temperature from the image signal of the output of the sensor 1, a human body area detector 3 for detecting the human body area in the image based on the output signal of the calculator 2 and the image signal, a person area divider 4 for dividing the human body area to the personal areas of the individual areas according to the output signal of the detector 3 and the image signal, and an image information detector 5 for outputting the detailed personal information of the person or the environmental information signal from the output signal of the divider 4 and the image signal.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The image processing apparatus of the present invention relates mainly to an image processing apparatus for detecting human information from a thermal image obtained from an infrared sensor. It can be used in many industrial fields such as Intelligent Building Systems (IBS).

[0002]

The field to be solved by the present invention relates to equipment control using a thermal image of an infrared sensor and using human information obtained therefrom. Conventionally, there has been a device that detects the position of a person and controls the air conditioning by using an infrared sensor. For example, the structure described in Japanese Patent Laid-Open No. 6-17883 is known. The configuration will be briefly described below. Infrared sensors measure the temperature of people and the environment in the detection area. The thermal image output from the infrared sensor is divided into blocks, and the cut-out temperature is determined from the average temperature of each block to detect the human body region. Neural nets or pattern recognition are used to output personal information signals such as the number of people and foot positions and environmental information signals according to the shape and size of the human body region. The comfort level of each individual is detected from the personal information signal and the environmental information signal to control the air conditioning equipment.

In addition to the air conditioners, accurate information about human beings is required in surveillance systems and the like. A method of detecting information about human beings is to install a visible surveillance camera for human monitoring or to use video. It is common to record. Although there are monitoring systems that use image processing, detection is often difficult due to problems such as illuminance and color. In addition, there are also surveillance systems that use infrared cameras, but they generally require a large number of detection elements or a cooling device to obtain a high-resolution thermal image, which requires complicated processing and high cost. Is required.

[0004]

However, the control device of the air conditioner using the infrared thermal image shown in the prior art corresponds to the average temperature of each fixed area where the human body and the background are not separated. Then, the human body region is detected within a predetermined cutting temperature range. Therefore, when the room temperature becomes high, the human body region may not be accurately detected. Further, when a plurality of people are close to each other, it is difficult to detect the number of people and to accurately detect the position of each person.

The first problem of the prior art is the problem that occurs when the temperature for cutting out the human body temperature from the thermal image of the infrared sensor is set in advance, regardless of the heat distribution in the external environment. The second problem is a problem caused by the low resolution of the image reading sensor. When the resolution of the image is low, it is difficult to accurately detect the number of people and the position of each person when a plurality of people are nearby.

The present invention solves such a conventional problem by accurately detecting the area of the human body in the detection area from the infrared image, and by detecting the area of the human body from the infrared image. An object of the present invention is to accurately divide each person into individual areas when the human body area is an overlap of a plurality of people.

[0007]

In order to achieve the above-mentioned object, a first technical solution in the present invention is, for an image processing apparatus, at least the temperature of a person and the environment in a detection area as a thermal image signal. A sensor unit for outputting, a background temperature calculation unit that processes a thermal image signal output from the sensor unit and calculates a background temperature from the thermal image signal, and an output signal and a thermal image signal of the background temperature calculation unit In addition, it is composed of a human body area detecting section for detecting a human body area in the thermal image and an image information detecting section for outputting a more detailed human personal information signal or environment information signal than the output signal of the human body area detecting section.

The second technical solution is constituted by adding a human area dividing section for dividing the human body area output from the human body area detecting section into individual areas which are areas of each person. ing.

[0009]

The image processing apparatus of the present invention has the following operations due to the above-mentioned configuration. As a first action, in the present invention, the temperature of the region considered to be the background is calculated from the thermal image signal output from the two-dimensional infrared sensor, and the cut-out temperature of the human body region is determined from this background temperature to obtain a more accurate human body. Area detection is possible.

As a second action, it is assumed that one human has one maximum in the human region dividing section, and one human body region in which a plurality of people exist based on the maximum of the thermal image signal. Is divided into individual areas for each person, the number of people can be accurately detected even when part of the human body overlaps.

The reason why one person has one maximum value is that when there is one person in the detection area of the infrared sensor, the exposed skin portion is more than the clothing portion. High temperature is detected. Also, of the exposed parts of the face or hands, the face part detects the highest temperature. This is because when an infrared sensor having a small number of pixels is used, the temperature of the face of each person on the thermal image has a maximum value on the thermal image.

[0012]

【Example】

(First Embodiment) A first embodiment of the image processing apparatus of the present invention will be described below with reference to the drawings. FIG. 1 is a block diagram of the image processing apparatus of this embodiment. Reference numeral 1 denotes a sensor unit that outputs a thermal image signal. Reference numeral 2 denotes a background temperature calculation unit that detects the background temperature signal from the thermal image signal. Reference numeral 3 denotes a human body area detection unit that detects a human body area. Reference numeral 4 denotes a human region dividing unit which divides the human body region into individual regions. An image information detection unit 5 estimates the number of people in the detection area and the amount of clothing from the thermal image.

Next, each part of FIG. 1 will be described. First, the sensor unit 1 will be described. The outside world 6 is an area in which a person exists, and the two-dimensional infrared sensor 14 included in the sensor unit 1 scans it to detect the temperature of an area that can be detected as thermal image information, and outputs a thermal image signal 8.

Next, the background temperature calculation unit 2 will be described. 1
Reference numeral 5 is a second-order differential calculating means for calculating a second-order differential of the thermal image signal 8 detected by the two-dimensional infrared sensor 14 and outputting a second-order differential image signal, and 16 is a second-order differential output of the second-order differential calculating means. The unevenness of the thermal image signal 8 is detected from the differential image signal, the temperature of the human region is set to 30 degrees, and when the temperature of the human region is higher than the indoor temperature, the temperature of the concave portion which is the background temperature is calculated. It is a background area temperature calculating means for calculating the temperature of the convex portion which is the background area and outputting the background temperature signal 9.

Next, the human body area detecting section 3 will be described. 1
Reference numeral 7 is a background temperature signal 9 which is an output of the background temperature calculation unit 2, and when the temperature Tman in the human region is higher than the room temperature Troom, the temperature obtained by adding the background temperature Tback and the threshold value θ is set as the cut-out temperature. If the temperature is low, the temperature obtained by subtracting the threshold value θ from the background temperature is set as the cutout temperature, and the cutout temperature determining means for outputting the cutout temperature signal, 18 is the cutout temperature signal determined by the cutout temperature determining means 17, and if Tman ≧ Troom For example, if Tback + θ <G Tman <Troom, the human body detection unit detects the pixel G obtained by Tback−θ> G as the human body region and outputs the human body region signal 10.

Next, the image information detecting section 5 will be described. 1
Reference numeral 9 is a person number detecting means for detecting the number of persons in the detection area from the personal area signal 11 output from the human area dividing unit 4 and outputting the number of people signal.

FIG. 2 is a block diagram showing a detailed internal structure of the human area dividing section 4. Reference numeral 20 is a maximum value detecting means for detecting a maximum value of the thermal image from the thermal image signal and outputting the maximum value signal, and 21 is a maximum value signal and a human body region signal output from the maximum value detecting means, and a maximum value to a human body. It is a maximum value personal area dividing unit that divides the human body area into individual areas according to the distance of pixels indicating the area and outputs a personal area signal.

Next, the operation of this embodiment configured as described above will be described. First, the overall operation will be described. The outside world 6 is an area where a human is present, and the two-dimensional infrared sensor 14 included in the sensor unit 1 scans the area to detect the temperature of an area that can be detected as thermal image information. The room may also be referred to as an area where people live. The heat distribution of the outside world 6 is 2
It is scanned by the three-dimensional infrared sensor 14 and input to the background temperature calculation unit 2 as the thermal image signal 8. Background temperature calculation unit 2
At, the background area is detected from the thermal image signal 8 which is the output of the sensor unit 1, and the background temperature is calculated. The thermal image signal 8 and the background temperature signal 9 that is the temperature of the background region of the thermal image signal 8 are input to the human body region detection unit 3. The human body region detection unit 3 detects a human body region based on the thermal image signal 8 and the background region signal 9, and outputs a human body region signal 10. Thermal image signal 8 in human region dividing section 4 and human body region signal 1 output in human body region detecting section 3
Based on 0, the human body region is divided into individual regions, which are regions for each person, and a personal region signal 11 is output. The thermal image signal 8, the background temperature signal 9 and the personal area signal 11 are input to the image information detection unit 5. In the image information detection unit 5, from the thermal image signal 8, the background temperature signal 9 and the personal area signal 11, the personal information signal 12 such as the number of people and the amount of clothes in the detection area, the background temperature and the position of the non-human heating element, etc. Environmental information signal 13
Is output.

Next, details of the operation of the above embodiment will be described. First, the operation of the background temperature calculation unit 2 will be described. Conventionally, in the case of detecting the background temperature from the thermal image signal 8, if there is no heat generation other than human beings in the room and the room temperature is 25 degrees or lower, by detecting the pixels of the temperature of 26 degrees to 34 degrees, The human body area is detected, and the background temperature is detected with the other area as the background. However, with the above configuration, the background temperature may not be accurately calculated from the two-dimensional thermal image. For example, the temperature distribution in the room may vary depending on the floor and each wall surface. There is also a method of detecting a person from the boundary line of the person without using the temperature range for cutting out the person, but accurate detection is difficult if a good edge cannot be detected.

Therefore, in this embodiment, a human being is regarded as a heat generating object, and when the temperature of the human region is lower than the ambient temperature, it is convex, and when the ambient temperature is high, it is concave. . The second-order differential calculating means 15 for calculating the second-order differential of the thermal image signal 8 finds irregularities in the thermal image signal. Assuming that the temperature of the human region is 30 degrees, the background region temperature calculation means 16 calculates the background temperature by setting the concave portion when the temperature is higher than the ambient temperature and the convex portion when the temperature is lower than the ambient temperature as the background region. .

Here, a polynomial fitting method, which is one of well-known moving average methods in image processing, is used as a method for capturing large irregularities in a thermal image. The image differential is obtained by changing the weighting factor of the polynomial fitting method, and at the same time the image is smoothed. The derivative of this polynomial fitting method is 2
By performing the operation twice, the second derivative of the thermal image is obtained while removing noise. From this second-order differential image signal, the method of finding the unevenness corresponding to a person on the thermal image and cutting out the region based on the temperature of the uneven part was used. In addition to the polynomial fitting method, it is also possible to obtain the second derivative of the thermal image by using a Laplacian operator of the four-neighbor type or the eight-neighbor type to capture the irregularities on the thermal image.

Next, the operation of the human body area detecting section 3 will be described. In the present embodiment, when the human area is higher than the indoor temperature, the thermal image signal 8 of a portion larger than the sum of the background temperature signal 9 and the threshold value θ is detected as the human body area from the background temperature signal 9, and the indoor temperature is detected. If it is lower, the thermal image signal 8 of a portion smaller than the background temperature signal 9 minus the threshold value θ is detected as the human body region.

Next, the operation of the human area dividing section 4 will be described. Thermal image signal 8 detected by the two-dimensional infrared sensor 14
Then, the personal area is detected from the human body area signal 10 output from the human body area detecting unit 3. Conventionally, when a plurality of persons are present in the detection area, one person is detected by the human body area detection unit 3.
If a person is not cut out, even if a plurality of humans exist in one human body area, only one person recognizes that they exist in that human body area. Therefore, in this embodiment, when a plurality of humans exist in one human body region, the number of humans is two.
One human body region is divided into individual regions assuming that it has one maximum value on the three-dimensional thermal image.

A detailed description of the operation of the human area dividing section 4 will be given with reference to FIG. 1 detected by the human body area detection unit 3
It is assumed that multiple humans exist in one human body area. Assuming that one person has one maximum value, the maximum value detecting means 20 calculates the maximum value from the thermal image signal 8 and outputs the maximum value signal. Assuming that the maximum value in the human body area is one person, the maximum value personal area division is performed based on the maximum value signal output by the maximum value detecting means 20 and the human body area signal output by the human body area detecting unit 3. The means 21 divides the personal area. If the distance of the human body region from the maximum value is short by using the distance on the thermal image by the maximum value personal region dividing means 21,
It is divided as the human personal area indicated by this maximum value. Here, when the aspect ratio of the thermal image is not 1, the distance is weighted using the aspect ratio of the thermal image, and one human body region is divided into individual regions by this weighting.

Next, the operation of the person number detecting means 19 will be described. Conventionally, when detecting the number of people in an image from an image, the area of a part that is more human than the image is detected, it is determined whether or not a person is a person based on the shape of that area, and the number of people is calculated based on how many there are. Was there. However, with this method, it is difficult to detect the number of people properly unless the shape of the region can be accurately recognized from the image. Also, when humans overlap, it becomes difficult to calculate the number of people. Therefore, in this embodiment, the number of persons is calculated by the human body area detection unit 3 and the human area division unit 4.

First, the human body region detecting section 3 detects the human body region. When a plurality of people are overlapped in the area detectable by the two-dimensional infrared sensor 14 and only one human body area can be detected by the human body area detecting unit 3, the operation of the human body area dividing unit 4 causes one person Is divided into personal areas. The number of personal areas indicates the number of people in the detection area.

(Second Embodiment) A second embodiment of the image processing apparatus of the present invention will be described below with reference to the drawings. In the first embodiment, the human body region dividing unit 4 is the sensor unit 1.
Of the thermal image signal 8 which is the output of the human body region signal 1 output from the human body region detection unit 3 and the maximum value signal.
From 0, one human body area was divided into individual areas.
In the present embodiment, the vertical and horizontal differentials of the thermal image signal 8 that is the output of the sensor unit 1 are taken to find the local minimum values in the vertical and horizontal directions, and the connection of these local minimum values is taken as one human body region. This is a boundary line that divides the personal area.

FIG. 3 is a block diagram showing a detailed internal structure of the human area dividing section 4. 22 is a minimum value detecting means for obtaining a minimum value in the vertical direction and the horizontal direction by differentiating the thermal image signal in the vertical and horizontal directions to output a minimum value signal; and 23 is a minimum value which is the output of the minimum value detecting means. Concatenate the value signals,
A minimum value connecting means for outputting a minimum value connecting signal; 24 divides the personal area from the human body area signal using the minimum value connecting signal output from the minimum value connecting means as a boundary line of the personal area;
It is a minimum value personal area dividing means for outputting a personal area signal.

A detailed description of the operation of the human body region dividing section 4 will be given with reference to FIG. 1 detected by the human body area detection unit 3
It is assumed that multiple humans exist in one human body area.
It is assumed that one person has one maximum value, and the boundaries of other humans have minimum values in the vertical and horizontal directions. Therefore, in this embodiment, the vertical and horizontal differentials are obtained to find the minimum value, and one human body region is divided into individual regions using the connection of these minimum values as a human boundary line.

In the human body region dividing section 4 of FIG. 3, the thermal image signal 8 is subjected to a vertical and horizontal differential operator, and the vertical and horizontal minimum values are detected by the minimum value detecting means 22. The minimum value signal output from the minimum value detecting means 22 is
Depending on the shape of the infrared image, it does not show continuous lines. Therefore, while paying attention to the aspect ratio of the thermal image, the minimum values are connected according to the distance between the pixels having the minimum value. The minimum value connection means 23, which receives the minimum value signal, connects the minimum value signals and outputs the minimum value connection signal as a boundary line that divides one human body region into one person region. Based on the boundary line of the personal area indicated by the minimum value connection signal and the human body area signal output from the human body area detection unit 3, the minimum value personal area dividing means 24 divides one human body area into individual areas.

The differential operator used here has a smoothing function P that takes the differential of the thermal image while removing noise.
They are a rewitt operator and a Sobel operator.

(Embodiment 3) Next, a monitoring apparatus will be shown as a specific embodiment using the image processing apparatus of the present invention.

In the conventional monitoring system, a person constantly monitors a visible image. This is because the technology for recognizing a person from a visible image has been developed, but few have been put to practical use. However, many techniques for detecting a moving object from a visible image have been put into practical use. However, it is difficult to accurately capture a person from a moving object because the objects captured by the movement include objects other than the person. In addition, it is impossible to detect a person who moves little. On the other hand, although it is easy for an infrared image to capture a person as a heat-generating object, a heat-generating object other than a person is also detected at the same time.

Therefore, in the present embodiment, the apparatus for detecting a moving object from visible image processing and the image processing apparatus of the present invention for detecting a heat-generating object from image processing using an infrared sensor are used together to detect human or environmental information. 2 shows a monitoring device with improved accuracy.

An embodiment of the present invention will be described below with reference to the drawings. FIG. 4 is a schematic block diagram of a monitoring device according to an embodiment of the present invention. Reference numeral 25 denotes a visible image sensor unit (camera) that outputs a visible image signal 30. Reference numeral 26 denotes a visible image processing unit, which converts a visible image signal 29 to a differential image signal 31.
Is output. A visible image information detection unit 27 outputs the moving object signal 32 from the differential image signal 31. Reference numeral 28 denotes a monitoring information detection unit, a moving object signal 31 output from the visible image information detection unit 27, a personal information signal 12 output from the image processing apparatus of the present invention, and an environment information signal 13 to a suspicious person warning signal 33 and an abnormal signal. 34 is output.

Here, it is assumed that the image information detecting section 5 additionally includes a foot position estimating means 35 and a posture determination estimating means 36. Each means estimates the foot position and posture by using a neural network or pattern recognition based on the human body region signal 10 or the personal region signal 11 which is the input of the image information detection unit 5. The output of each means is added to the personal information signal 12 which is the output of the image information detection unit 5 and output.

Next, the above-mentioned embodiment constructed as described above will be explained. The visible light 29 from the outside world 6 by the camera 25
To output a visible image signal 30. The difference image signal 31 is output by taking the temporally continuous difference of the visible image signal 30 in the visible image processing unit 26. In the visible image information detection unit 27, the presence / absence of a moving object, the number of moving objects, and the position of the moving object are calculated from the difference image signal 31 as a moving object signal 32.
Output as. In the monitoring image information detection unit 28, the moving object signal 32 and the personal information signal 12 and the environmental information signal 13 output from the image processing apparatus of the present invention correspond to the moving object and the heat-generating object in the number and position in the detection area. Check whether or not it is taken and detect the human body. The position where the human body is detected is the off-limit area, and when the position of the human body does not change, the suspicious person warning signal 33 and the abnormal signal 34 are output.
Further, if there is a region having a temperature much higher than the temperature of the human body in the detection area, the abnormal signal 34 is output because there is a possibility of fire.

(Embodiment 4) Next, a personal recognition apparatus will be shown as a specific embodiment using the image processing apparatus of the present invention. Hereinafter, the present embodiment will be described with reference to the drawings. FIG. 5 is a schematic block diagram of the personal identification device in this embodiment. FIG. 5 is a schematic block diagram of the third embodiment. The monitoring information detection unit 28 shown in FIG. 4 is replaced with the personal identification unit 37, the visible image signal 30 output from the camera unit 25 is added to the input, and the personal identification signal is added. 38 is output. Further, the personal identification section 37 includes a personal information storage means 39 and a personal recognition means 4
Has 0.

The operation of the personal identification section 37 will be described. The personal information storage means 39 stores the position, posture, physique, color of clothes, etc. of the person. The personal recognition means 40 is the personal information storage means 3
Based on the characteristics of the individual stored in 9, the individual can be identified or tracked by a neural network or pattern recognition. This makes it possible to identify who is in the room. There are applications from this personal recognition device to a personal audience rating measuring device.

(Embodiment 5) Next, a nursing apparatus will be shown as a specific embodiment using the image processing apparatus of the present invention.

The present embodiment will be described below with reference to the drawings. FIG. 6 is a schematic block diagram of the nursing device in this embodiment. Reference numeral 41 is a non-nurse state determination unit, which is an output of the image information detection unit 5 and is a personal information signal 12 and an environmental information signal
And outputs a non-nurse status signal 42 based on Here, it is assumed that the image information detection unit 5 is additionally provided with a posture determination estimation unit 36, a clothing amount estimation unit 43, and a skin temperature estimation unit 44. Each means is based on the human body region signal 10 or the personal region signal 11 which is the input of the image information detection unit 5,
Using neural networks or pattern recognition
Estimate posture, amount of clothing, and skin temperature. The output of each means is added to the personal information signal 12 which is the output of the image information detection unit 5 and output.

The operation of the non-nurse status determination section 41 will be described. The posture of the non-nurse is discriminated from the personal information signal 12 output from the image information detection unit 5, and whether the non-nurse is in bed, awake, awaking, sleeping, etc. Output status. In addition, the state such as whether the bedding does not deviate from the clothing amount or the skin temperature or whether the body temperature is higher than usual is output. The environment information signal 13 outputs the indoor condition such as whether the room temperature is not uncomfortable for the non-nurse or whether air conditioning is required. Since this nursing device outputs the status of non-nurses and the situation inside the room,
You can care even if the caregiver is not around for 24 hours.

(Embodiment 6) Next, an air conditioner will be shown as a specific embodiment using the image processing apparatus of the present invention. Hereinafter, the present embodiment will be described with reference to the drawings. FIG. 7 is a schematic block diagram of the air conditioner in this embodiment.
A room temperature sensor 45 outputs the room temperature as a room temperature signal 51. Reference numeral 46 is a humidity sensor, which indicates the indoor humidity by a humidity signal 52.
Output as. An outdoor temperature sensor 47 outputs the outdoor temperature as an outdoor temperature signal 53. Reference numeral 48 denotes wind direction / volume information, which is the wind direction / volume signal 5 of the wind emitted from the outlet of the air conditioner.
4 is output. Reference numeral 49 is a control index determination unit, which is the output of the image processing apparatus of the present invention and is the personal information signal 12 and the environmental information signal 1.
3, the room temperature signal 51 that is the output of the room temperature sensor 45, the humidity signal 52 that is the output of the humidity sensor 46, the outside air temperature signal 53 that is the output of the outside air temperature sensor 47, and the output of the wind direction / air volume information 48. The wind direction / air volume signal 54 is input, and the activity amount and comfort level of each person in the room are output as a control signal 55. Reference numeral 50 is an air-conditioning control unit, which is the control index determination unit 49.
The air conditioner is controlled from the control signal 55 which is the output of the.

The control index determination unit 49 will be described. In the personal information signal 12 input to the control index determination unit 49 by the activity amount detection means 56, the activity amount is calculated from the movement amount obtained from the change in the individual foot position and the change in posture such as standing, sitting or lying down. To detect. Reference numeral 57 is a comfort level detecting means, and the personal information signal 12 and the environmental information signal 1 are input to the control index determining section 49.
3, indoor temperature signal 51, humidity signal 52, and outside air temperature signal 53
And PMV (ISO 7
730-1984) and the like, and outputs a control signal 55 corresponding to the comfort level to the air conditioning controller 50.

Here, it is assumed that the image information detecting section 5 has a structure of a foot position estimating means 35, a posture estimating means 36, and a clothing amount estimating means 43. Based on the human body region signal 10 or the personal region signal 11 input to the image information detection unit 5, each means estimates the foot position, posture, and amount of clothing by using a neural network or pattern recognition. The output of each means is added to the personal information signal 12 which is the output of the image information detection unit 5 and output.

The operation of the activity amount detecting means 56 will be described.
When PMV is used as a control index as the degree of comfort, it is known that the amount of activity is an important factor as a human state.
In order to detect the amount of activity, the personal information signal 12 output from the image information detection unit 5 is used to store the foot position and posture of each person in the room, and the movement of each person in the room is detected over time. . If there is a movement, the movement amount is associated with the moving speed, and if there is no movement, the activity amount is detected from the difference in posture, and the activity amount is detected.

The operation of the comfort level detecting means 57 will be described.
When calculating PMV, which is one of the indicators of comfort level, air temperature, radiation temperature, humidity, air flow, metabolic rate, and clothing amount are required, but the metabolic rate and clothing amount, which are physical quantities related to humans, must be measured. It is difficult. Therefore, in this embodiment, the amount of clothing is estimated by the environment information signal 13, the room temperature signal 51, the humidity signal 52, the wind direction / air volume signal 54, and the image information detection unit 5,
The activity level detection means 56 estimates the metabolic rate from the activity level to calculate the comfort level PMV for each individual. Since the accurate information on the number of people is output from the personal information signal 12 from the image processing apparatus, more accurate internal heat generation by the human body in the room can be detected, and the correction according to the heat load in the room can be added.
A PMV that is a representative of the room is calculated from this correction and is output to the air conditioning control unit 50.

In the present embodiment, the comfort level of each individual is calculated from the personal information signal 12, the environment information signal 13, the room temperature signal 51, the humidity signal 52, the outside air temperature signal 53, and the wind direction / air volume information 54 which are the outputs of the image processing apparatus. It is possible to improve the comfort level of an individual by detecting the detected comfort level that is representative of the room and controlling the room temperature, air volume, and wind direction. In addition, since an image processing device having a small number of pixels is used, the cost is low and it is easy to use.

The image processing apparatus of the present invention can be used in various other devices.

[0050]

As described above, according to the present invention, the human body region detecting section accurately determines the region of the human body in the detection area from the infrared image by determining the cutting temperature of the human body region based on the background temperature of the infrared image. Can be detected.

Further, when a human body region is detected from an infrared image and one human body region overlaps a plurality of humans, one human body region is divided into 1 by the human region dividing unit.
It can be divided into the exact personal areas of one person.

As described above, according to the present invention, since the human area in the detection area and each person can be accurately detected from the infrared image, the labor of the person can be reduced in the monitoring device and the nursing device using the same. Therefore, in an air conditioner or the like, control suitable for the situation of the person in the room becomes possible.

[Brief description of drawings]

FIG. 1 is a configuration diagram of an image processing apparatus according to a first embodiment of the present invention.

FIG. 2 is a configuration diagram of a human area dividing unit in the same embodiment.

FIG. 3 is a configuration diagram of a human area dividing unit according to a second embodiment of the present invention.

FIG. 4 is a configuration diagram of a monitoring device according to a third embodiment of the present invention.

FIG. 5 is a configuration diagram of a personal recognition device according to a fourth embodiment of the present invention.

FIG. 6 is a configuration diagram of a nursing device according to a fifth embodiment of the present invention.

FIG. 7 is a configuration diagram of a control device for an air conditioner according to a sixth embodiment of the present invention.

[Explanation of symbols]

 DESCRIPTION OF SYMBOLS 1 Sensor part 2 Background temperature calculation part 3 Human body region detection part 4 Human region division part 5 Image information detection part 6 Outside world 7 Thermal distribution of outside world 8 Thermal image signal 9 Background temperature signal 10 Human body region signal 11 Personal area signal 12 Personal information signal 13 environment information signal 14 two-dimensional infrared sensor 15 second-order differential calculation means 16 background area temperature calculation means 17 cutout temperature determination means 18 human detection means 19 human number detection means 20 maximum value detection means 21 maximum value area dividing means 22 minimum value detection means 23 minimum value connecting means 24 minimum value region dividing means 25 camera 26 visible image processing unit 27 visible image information detecting unit 28 monitoring information detecting unit 29 visible light 30 visible image signal 31 differential image signal 32 moving object signal 33 suspicious person warning signal 34 Abnormal signal 35 Foot position estimation means 36 Posture determination estimation means 37 Personal identification section 3 Individual identification signal 39 Personal information storage means 40 Individual recognition means 41 Non-nurse status determination unit 42 Non-nurse status signal 43 Clothes amount estimation means 44 Skin temperature estimation means 45 Room temperature sensor 46 Humidity sensor 47 Outside temperature sensor 48 Wind direction / air volume information 49 Control Index Determining Section 50 Air Conditioning Controller 51 Room Temperature Signal 52 Humidity Signal 53 Outside Air Temperature Signal 54 Wind Direction / Air Volume Signal 55 Control Signal 56 Activity Level Detection Means 57 Comfort Level Detection Means

─────────────────────────────────────────────────── ─── Continuation of the front page (51) Int.Cl. ⁶ Identification code Internal reference number FI Technical display location G01D 21/02 G01J 1/02 W 9309-2G G01V 8/10 8/12 G06M 11/00 D ( 72) Inventor Nakamoto Son 3-10-1 Higashisanda, Tama-ku, Kawasaki City, Kanagawa Prefecture Matsushita Giken Co., Ltd.

Claims (12)

[Claims] 1. A sensor unit for measuring the temperature of a human being and an environment in a detection area and outputting an infrared image (hereinafter referred to as a thermal image signal), and a thermal image signal output from the sensor unit to detect a background region. A background temperature calculation unit that calculates the temperature and outputs a background temperature signal, a background temperature signal that is the output of the background temperature calculation unit, and an output of the sensor unit based on the thermal image signal that is the output of the sensor unit. A human body region detection unit that detects a human body region from the thermal image signal and outputs a human body region signal, a human body region signal that is the output of the human body region detection unit, and a thermal image signal that is the output of the sensor unit. An image processing apparatus comprising an image information detection unit that outputs a human personal information signal or an environment information signal based on the background temperature signal output from the background temperature calculation unit. 2. The background temperature calculation unit uses a second-order differential calculation unit that performs second-order differentiation of the thermal image signal output from the sensor unit and outputs a second-order differential image signal, and an output of the second-order differential calculation unit. Based on a certain second-order differential image signal and the thermal image signal output from the sensor unit, the region that is the concave (or convex) portion of the thermal image signal is considered as the background, the average temperature is calculated, and the background temperature signal is calculated. The image processing apparatus according to claim 1, further comprising a background area temperature calculation unit that outputs the background area temperature. 3. A human body region detection unit determines a human body cutout temperature based on a background temperature signal output from a background temperature calculation unit, and outputs a cutout temperature signal, and a cutout temperature determination unit, and the cutout temperature determination unit. 3. A human body detecting means for detecting a human body region based on a cutout temperature signal which is an output and a thermal image signal which is an output of the sensor unit, and outputting a human body region signal. Image processing device. 4. A human region dividing unit that divides a human body region into individual regions based on a human body region signal output from a human body region detecting unit and a thermal image signal output from a sensor unit, and outputs a personal region signal. Based on the personal area signal which is the output of the human area dividing section, the thermal image signal which is the output of the sensor section, and the background temperature signal which is the output of the background temperature calculation section, at least the number of people in the detection area 2. An image information detecting section for outputting as an information signal is added.
The image processing device according to any one of 1 to 3. 5. A human region dividing unit calculates a maximum value from a thermal image signal output from a sensor unit and outputs a maximum value signal, and a maximum value output from the maximum value detecting unit. 5. The image processing apparatus according to claim 4, further comprising local maximum value personal area dividing means for dividing the human area based on the signal and the human body area signal output from the human body area detecting unit and outputting the individual area signal. . 6. A human region dividing unit detects a minimum value from a thermal image signal output from a sensor unit and outputs a minimum value signal, and a minimum value output from the minimum value detecting unit. A human body region is connected based on the minimum value connection means for connecting the signals and outputting the minimum value connection signal, the minimum value connection signal which is the output of the minimum value connection means, and the human body region signal which is the output of the human body region detection unit. 5. A minimum value personal area dividing means for dividing and outputting a personal area signal is provided.
The image processing device described. 7. The image information detecting section has a person number detecting means for detecting the number of persons in the detection area based on the personal area signal output from the human area dividing section and outputting the number of people signal. Item 7. The image processing device according to any one of items 1 to 6. 8. A monitoring device comprising the image processing device according to claim 1. 9. An image processing apparatus according to any one of claims 1 to 7, further comprising a personal identification section for learning a characteristic amount of a human obtained from the image processing apparatus and for identifying an individual. Personal identification device. 10. A nursing device comprising the image processing device according to claim 1. 11. An air conditioner control device comprising the image processing device according to claim 1. Description: 12. An applied device using the image processing device according to claim 1.