JPWO2021019768A1 - Indoor layout estimation device, indoor layout estimation method and air conditioner - Google Patents

Abstract

As the evaluation values of each of the plurality of pixels included in the image output from the infrared camera (1) capturing the target area, the temporal transition of each pixel value in the plurality of pixels and the outside temperature The evaluation value calculated by the similarity calculation unit (12) among the plurality of pixels included in the image and the similarity calculation unit (12) for calculating the similarity with the temporal transition is the first threshold value. The indoor layout estimation device (2) is configured to include a window area determination unit (13) for determining that a region including a larger pixel is a region where a window exists in an image.

Description

The present invention relates to an indoor layout estimation device, an indoor layout estimation method, and an air conditioner for determining an area where a window exists in a target area.

In the following Patent Document 1, a window that compares the luminance and luminance threshold of each of a plurality of pixels included in the image information and detects a region including pixels having a luminance larger than the luminance threshold as a window region. The detection method is disclosed.
In the window detection method disclosed in Patent Document 1, for example, when the entire air-conditioned room is brightly illuminated by sunlight, the brightness threshold value is set high so that the wall surface of the air-conditioned room and the window can be distinguished. NS. Further, for example, when the window is covered with a curtain and the entire air-conditioned room is dark, the brightness threshold value is set low.

Japanese Unexamined Patent Publication No. 2017-44373

The brightness of a plurality of pixels included in the image information varies depending on the outside air temperature. However, in the window detection method disclosed in Patent Document 1, since the luminance threshold value is set regardless of the outside air temperature, the window area is also detected regardless of the outside air temperature. Therefore, the window detection method has a problem that erroneous detection of the window area may occur when the outside air temperature changes.

The present invention has been made to solve the above-mentioned problems, and an object of the present invention is to obtain an indoor layout estimation device capable of preventing erroneous detection of a window area even if the outside air temperature changes.

In the indoor layout estimation device according to the present invention, as the evaluation value of each of the plurality of pixels included in the image output from the infrared camera capturing the target area, the temporal time of each pixel value in the plurality of pixels is used. The first is the evaluation value calculated by the similarity calculation unit among the plurality of pixels included in the image and the similarity calculation unit that calculates the similarity between the transition and the temporal transition of the outside temperature. A region including pixels larger than the threshold value is provided with a window region determination unit for determining that the region includes a window in the image.

According to the present invention, it is possible to prevent erroneous detection of the window region even if the outside air temperature changes.

It is a block diagram which shows the air conditioner which includes the room layout estimation apparatus 2 which concerns on Embodiment 1. FIG. It is a block diagram which shows the room layout estimation apparatus 2 which concerns on Embodiment 1. FIG. It is a hardware block diagram which shows the hardware of the room layout estimation apparatus 2 which concerns on Embodiment 1. FIG. It is a block diagram which shows the inside of the standard deviation calculation part 14 which concerns on Embodiment 1. FIG. It is a hardware block diagram of the computer when the room layout estimation apparatus 2 is realized by software, firmware and the like. It is a flowchart which shows the room layout estimation method which is the processing procedure of the room layout estimation apparatus 2. Of the plurality of pixels p _{1 to p N} included in each of the images G _{m for} the past 12 hours, the pixel values pd _{n of one pixel pn, m} included in the window area W _m, It is explanatory drawing which shows an example of the temporal transition of _m. Is an explanatory diagram showing an example of calculation results of evaluation value S ₁ to S _N by the similarity calculation unit 12. It is explanatory drawing which shows an example of the determination result of the _{window area W m} by the window area determination unit 13. It is explanatory drawing which shows an example of a certain image G _{m among M} image capture images G _{1 to} GM. It is explanatory drawing which shows the background area B _m and the foreground area F _m in a certain image G _{m among M} image capture images G _{1 to} GM. Is an explanatory view showing the lower end of the pixels in the foreground area _F 1 to F _M in the M image _G 1 ~G _M. It is explanatory drawing which shows an example of the determination result of the _{floor area Y m} by the floor area determination unit 16. It is explanatory drawing which shows the boundary line between a floor surface and a wall surface, a floor area Y _m , and a window area W _m. It is a block diagram which shows the room layout estimation apparatus 2 which concerns on Embodiment 3. It is a hardware block diagram which shows the hardware of the room layout estimation apparatus 2 which concerns on Embodiment 3. FIG.

Hereinafter, in order to explain the present invention in more detail, embodiments for carrying out the present invention will be described with reference to the accompanying drawings.

Embodiment 1.
FIG. 1 is a configuration diagram showing an air conditioner including the indoor layout estimation device 2 according to the first embodiment.
In Figure 1, the infrared camera 1, at regular time intervals Delta] t, imaged repeatedly target area, an image obtained by imaging an object region, the indoor layout estimated image G ₁ ~G _M of M where the imaging time t are different from each other Output to device 2. M is an integer of 2 or more.
The constant time interval Δt is, for example, a time interval that enables imaging at any frame rate of 10 frames per second to 60 frames per second.

The target area is a part or the whole of a room where the air is adjusted by the air conditioner 4 described later, and the target area includes windows, floors, walls, human beings, desks, personal computers, and the like.
Target region, because it is fixed, M-number of the image G ₁ ~G _M is an image the same target area is imaged.
The image _Gm (m = 1, ..., M) is a direction parallel to the x-axis (hereinafter referred to as "x-direction") and a direction parallel to the y-axis (hereinafter referred to as "y-direction"). It is an image of a two-dimensional plane represented by. E.g., x-direction of the image G _m corresponds to any horizontal space target area exists, y direction of the image G _m corresponds to the vertical direction of the space region of interest is present. In addition, in this specification, "parallel" is not limited to parallel in a strict sense, but includes substantially parallel.
The image G _m includes N pixels p _{1, m to p N, m} , and the respective pixel values pd _{1, m to pd N, m in the pixels p 1, m to p N, m} are temperatures. Is shown. N is an integer of 2 or more.
For example, the pixel _p 1, 1 _{~p 1, M} contained in each of the M image _G 1 ~G _M has the same pixel pixel position to each other, ie, each of the image _G 1 ~G _M The pixels p _{2,1 to p 2, M} that are included are pixels having the same pixel position. Similarly, for example, an image _G 1 ~G pixel _p N included in each of the _M, 1 _{~p N, M} is the pixel positions are the same pixels from each other.

Room layout estimation device 2, based on the M image G ₁ ~G _M output from the infrared camera 1, in the target region, a region in which the window is located (hereinafter, referred to as "window area"), the floor Each of the area where the wall is located (hereinafter referred to as "floor area") and the area where the wall is located (hereinafter referred to as "wall area") is determined.
The indoor layout estimation device 2 outputs each of the determination result of the window area, the determination result of the floor area, and the determination result of the wall area to the control unit 3.

The control unit 3 adjusts the air in the target area by controlling, for example, the air conditioner 4 based on the determination result output from the indoor layout estimation device 2.
The air conditioner 4 is a device that adjusts the air in the target area.
In the air conditioner shown in FIG. 1, the control unit 3 controls the air conditioner 4 to adjust the air in the target region. However, this is only an example, and the control unit 3 may adjust the air in the target area by controlling a device such as a fan, a floor heater, or a panel heater.

FIG. 2 is a configuration diagram showing an indoor layout estimation device 2 according to the first embodiment.
FIG. 3 is a hardware configuration diagram showing the hardware of the indoor layout estimation device 2 according to the first embodiment.
The indoor layout estimation device 2 includes an image storage unit 10, an outside air temperature storage unit 11, a similarity calculation unit 12, a window area determination unit 13, a standard deviation calculation unit 14, a foreground area determination unit 15, a floor area determination unit 16, and a wall area. A determination unit 17 is provided.

The image storage unit 10 is realized by, for example, the image storage circuit 20 shown in FIG.
Image storage unit 10 stores the M image _G 1 ~G _M output from the infrared camera 1.
The outside air temperature storage unit 11 is realized by, for example, the outside air temperature storage circuit 21 shown in FIG.
Outside temperature storage unit 11 is, for example, by a thermometer, was measured at constant time intervals Delta] t ', the measurement time t stores M different outside air temperature T ₁ through T _M together.
The outside air temperature T _m (m = 1, ..., M) is the air temperature outside the building including the target area, and a window exists between the target area and the outside of the building.
In the indoor layout estimation apparatus 2 shown in FIG. 2, the outside air temperature storage unit 11 stores the outside air temperature T ₁ through T _M measured by the thermometer. However, this is only an example, and the communication unit (not shown) of the indoor layout estimation device 2 downloads the temperature data near the building including the target area from, for example, a site on the Internet that distributes meteorological data. , outside temperature storage unit 11 may be configured to store the data of the temperature as the outside air temperature T ₁ through T _M.

Time interval Delta] t 'of the measurement in the M outside air temperature _T 1 through T _M, the time interval Delta] t of the imaging in the M image _G 1 ~G _M, are identical. Therefore, the measurement time t'of the _mth outside air temperature Tm _{and the imaging time t of the mth} image Gm are the same time.
However, the measurement time interval Δt'and the imaging time interval Δt are not limited to those that are exactly the same, and the measurement time interval Δt'and the imaging time interval Δt are within a range where there is no practical problem. May be different from each other. Further, the m-th measurement time t'and the m-th imaging time t are not limited to those that are exactly the same, and the m-th measurement time t'and m are within a range where there is no practical problem. The second imaging time t may be different from each other.

The similarity calculation unit 12 is realized by, for example, the similarity calculation circuit 22 shown in FIG.
Similarity calculation unit 12 obtains the M pieces of the image _G 1 ~G _M stored in the image storage unit 10.
The similarity calculation unit 12 has N pixels p _{1, m to} p as evaluation values S _{1 to} NS of _N pixels p _{1, m to p N, m included in the image G m.} The degree of similarity between the temporal transition of each pixel value pd _{1, m to pd N, m in N and m and} the temporal transition of the outside temperature T _{m stored in the outside temperature storage unit 11 is calculated.} ..
Similarity calculation unit 12 outputs the respective evaluation values S ₁ to S _N of calculated in the window region judgment section 13.

The window area determination unit 13 is realized by, for example, the window area determination circuit 23 shown in FIG.
Window region judgment section 13, of the M image G ₁ ~G _M stored in the image storage unit 10, and acquires any one of the images G _m.
The window area determination unit 13 has _{an evaluation value Sn} (n = 1) calculated by the similarity calculation unit 12 among the N pixels p _{1, m to p N, m included in the acquired image G m.} , ..., N) is determined to be the window region W _{m in the image G m} , which includes pixels larger than _{the first threshold value Th 1.} The higher the similarity between the temporal changes of the outside air temperature T _m, there is a high possibility of the window area W _m, the region evaluation value S _n comprises greater pixel than the first threshold value Th _1, the image G it is determined to be a window region W _m in the _m. The first threshold value Th ₁ may be stored in, for example, the internal memory of the window area determination unit 13 or may be given from the outside of the indoor layout estimation device 2 shown in FIG.
As a determination result of the _{window area W m} , the window area determination unit 13 _{inputs, for example, data indicating coordinates in the image G m in which the window area W m} exists to each of the control unit 3 and the wall area determination unit 17. Output.

As shown in FIG. 4, the standard deviation calculation unit 14 includes a pixel group specifying unit 14a and a deviation calculation processing unit 14b.
The standard deviation calculation unit 14 is realized by, for example, the standard deviation calculation circuit 24 shown in FIG.
Standard deviation calculating unit 14 obtains the M pieces of the image _G 1 ~G _M stored in the image storage unit 10.
The standard deviation calculation unit 14 _{includes pixels p 1} , 1 to p _{N, 1 , p 1 , 2 to} p N, 2, ..., P ₁ included in each of the _{M images G 1 to} GM. _, M _{~p N,} in the _M, N pieces of the pixel group pixel positions containing the same pixel (1) is calculated respective evaluation values _v 1 to v _N of ~ (N).
The pixel group (1) includes pixels p ₁ , 1, p ₁ , 2, ..., P _{1, M} , and the pixel group (2) includes pixels p ₂ , 1, p ₂ , 2, ...・ ・ Includes p _{2, M.} Further, the pixel group (N) includes pixels p _{N, 1} , p _{N, 2} , ..., P _{N, M.}
Standard deviation calculating section 14, a pixel group (n) (n = 1, ···, N) as the evaluation value _{v n} of, for example, M number of pixel _{p n, 1} which is included in the pixel group (n) Calculate the standard deviation of the pixel values pd _{n, 1 to pd n, M} in ~ _{pn, M.}
The standard deviation calculation unit 14 outputs the calculated evaluation values v _{1 to v N} to the foreground region determination unit 15.

FIG. 4 is a configuration diagram showing the inside of the standard deviation calculation unit 14 according to the first embodiment.
Pixel group identifying unit 14a obtains the M pieces of the image _G 1 ~G _M stored in the image storage unit 10.
The pixel group specifying unit 14a _{includes pixels p 1} , 1 to p _{N, 1 , p 1 , 2 to} p N, 2, ..., P ₁ included in each of the _{M images G 1 to} GM. _{, M to} p _{N, M} , each of N pixel groups (1) to (N) including pixels having the same pixel position is specified.
The deviation calculation processing unit 14b is included in the respective pixel groups (1) to (N) as the evaluation values v _{1 to v N} of the respective pixel groups (1) to (N) specified by the pixel group identification unit 14a. It is to have M pieces of pixel _{p n,} 1 _~p n, pixel values in the _{M pd n,} 1 _~pd n, calculates the standard deviation of _M.
The deviation calculation processing unit 14b outputs the calculated evaluation values v _{1 to v N} to the foreground region determination unit 15.

The foreground region determination unit 15 is realized by, for example, the foreground region determination circuit 25 shown in FIG.
In the foreground region determination unit 15, the evaluation value v _n (n = 1, ..., N) calculated by the standard deviation calculation unit 14 is the second among the N pixel groups (1) to (N). A pixel group (n) larger than the threshold value Th _{2 of is specified.}
The second threshold value Th ₂ may be stored in the internal memory of the foreground region determination unit 15, or may be given from the outside of the indoor layout estimation device 2 shown in FIG. 2, for example.
Foreground area determination section 15, each image _G N pixels _p 1 contained in the _m, m _{~p N,} of _m, the pixel _{p n,} which is included in the identified group of pixels _{(n), m} It is determined that the region including is the foreground region F _m in each image G _m.
Foreground region determining unit 15, as the determination result of the foreground region F _m, for example, and outputs the foreground region F _m is present, the data indicating the coordinates within the image G _m in the floor area determination unit 16.
The foreground area is an area where a moving object such as a human or an animal exists.

The foreground area F _m is an area other than the background area B _m in the image G _m. Therefore, the foreground region determination unit 15 may first determine the background region B _m and then determine the foreground region F _m . In this case, the foreground area determination unit 15 first among the N pixel groups (1) ~ (N), evaluation value v _n calculated by the standard deviation calculating section 14 of the second threshold value Th ₂ or less The pixel group (n) is specified. The foreground region determining section 15, each pixel _p 1 contained in the captured image _{G m,} m _{~p N,} of _m, the pixel _{p n,} which is included in the identified group of pixels _{(n), m} It is determined that the region including is the background region B _m in _{each captured image G m.} The foreground region determining unit 15, in the captured image G _m, the area other than the background area B _m may be determined as a foreground region F _m.

The floor area determination unit 16 is realized by, for example, the floor area determination circuit 26 shown in FIG.
The floor area determination unit 16 exists at the lower end position in the y direction of the image G _m among the plurality of pixels _{pn and m included in each foreground area F m} determined by the foreground area determination unit 15. The pixel is specified as a floor area pixel. If the foreground area F _m is, for example, an area where _{a human is present, the position of the lower end in the foreground area F m} corresponds to the position of the human foot, and the position of the human foot corresponds to the position of the floor. ..
The floor area determination unit 16 determines that the area including all the floor area pixels included in _{each foreground area F m} is the floor area Y _m in _{each image G m.}
Of the plurality of pixels p _{n, m} included in the foreground region F _m , the pixel existing at the lower end position in the y direction of the image G _m is the plurality of pixels p included in the _{foreground region F m. Of n and m} , among a plurality of pixels having the same x-coordinate, the pixel has the smallest y-coordinate.
As a determination result of the _{floor area Y m} , the floor area determination unit 16 _{inputs, for example, data indicating coordinates in the image G m in which the floor area Y m} exists to each of the control unit 3 and the wall area determination unit 17. Output.

The wall area determination unit 17 is realized by, for example, the wall area determination circuit 27 shown in FIG.
Wall area determining unit 17, and the window area W _m determined by the window region judgment section 13, on the basis of the a floor area Y _m determined by the floor area determination unit 16, a wall area K _m in the target area judge. The method for determining the wall area K _m will be described later.
Wall area determining unit 17, as the determination result of the wall area K _m, for example, outputs wall area K _m is present, the data indicating the coordinates within the image G _m to the control unit 3.

In FIG. 2, the image storage unit 10, the outside air temperature storage unit 11, the similarity calculation unit 12, the window area determination unit 13, the standard deviation calculation unit 14, the foreground area determination unit 15, and the floor, which are the components of the indoor layout estimation device 2, are shown. It is assumed that each of the area determination unit 16 and the wall area determination unit 17 is realized by dedicated hardware as shown in FIG. That is, the indoor layout estimation device 2 includes an image storage circuit 20, an outside air temperature storage circuit 21, a similarity calculation circuit 22, a window area determination circuit 23, a standard deviation calculation circuit 24, a foreground area determination circuit 25, a floor area determination circuit 26, and the like. It is assumed that it is realized by the wall area determination circuit 27.

Here, each of the image storage circuit 20 and the outside temperature storage circuit 21 is, for example, a RAM (Random Access Memory), a ROM (Read Only Memory), a flash memory, an EPROM (Erasable Programmable Read Only Memory), and an EPROM (Epil). It corresponds to a non-volatile or volatile semiconductor memory such as Read Only Memory), a magnetic disk, a flexible disk, an optical disk, a compact disk, a mini disk, or a DVD (Digital Versaille Disc).
Further, each of the similarity calculation circuit 22, the window area determination circuit 23, the standard deviation calculation circuit 24, the foreground area determination circuit 25, the floor area determination circuit 26 and the wall area determination circuit 27 is, for example, a single circuit or a composite circuit. A programmed processor, a parallel programmed processor, an ASIC (Application Specific Integrated Circuit), an FPGA (Field-Programmable Gate Array), or a combination thereof is applicable.

The components of the indoor layout estimation device 2 are not limited to those realized by dedicated hardware, but the indoor layout estimation device 2 is realized by software, firmware, or a combination of software and firmware. There may be.
The software or firmware is stored as a program in the memory of the computer. A computer means hardware that executes a program, and corresponds to, for example, a CPU (Central Processing Unit), a central processing unit, a processing unit, a computing device, a microprocessor, a microcomputer, a processor, or a DSP (Digital Signal Processor). do.
FIG. 5 is a hardware configuration diagram of a computer when the indoor layout estimation device 2 is realized by software, firmware, or the like.
When the indoor layout estimation device 2 is realized by software, firmware, or the like, each of the image storage unit 10 and the outside air temperature storage unit 11 is configured on the memory 31 of the computer. A program for causing a computer to execute the processing procedures of the similarity calculation unit 12, the window area determination unit 13, the standard deviation calculation unit 14, the foreground area determination unit 15, the floor area determination unit 16, and the wall area determination unit 17 is stored in the memory 31. Stored. Then, the processor 32 of the computer executes the program stored in the memory 31.

Further, FIG. 3 shows an example in which each of the components of the indoor layout estimation device 2 is realized by dedicated hardware, and FIG. 5 shows an example in which the indoor layout estimation device 2 is realized by software, firmware, or the like. Shows. However, this is only an example, and some components in the indoor layout estimation device 2 may be realized by dedicated hardware, and the remaining components may be realized by software, firmware, or the like.

Next, the operation of the air conditioner shown in FIG. 1 will be described.
Infrared camera 1, at regular time intervals Delta] t, imaged repeatedly target area as the image of the target region, and outputs the image G ₁ ~G _M of M where the imaging time t are different from each other in the room layout estimation device 2.
Image G ₁ ~G _M of the M output from the infrared camera 1 is stored in the image storage unit 10 of the indoor layout estimation device 2. The image storage unit 10, for example, images G _m for the past 24 hours are stored.

FIG. 6 is a flowchart showing an indoor layout estimation method which is a processing procedure of the indoor layout estimation device 2.
Similarity calculation unit 12, from the image storage unit 10, the imaging time t to obtain the M different image _G 1 ~G _M together.
If the current time, for example 00 pm 19, similarity calculation unit 12, from the image storage unit 10, starting from the 00 pm 19, for example, acquires image G _m for the past 12 hours.
FIG. 7 shows a pixel p included in the window area W _{m among the N pixels p 1, m to p N, m included in each of the images G m for the past 12 hours. n,} is an explanatory diagram showing an example of temporal transition of pixel values pd _{n, m} of the _m.

Also, the similarity calculating unit 12, from the outside temperature storage unit 11, the measurement time t 'to obtain the M different outside air temperature T ₁ through T _M together.
If the current time is 00 minutes at 19, the similarity calculation unit 12, from the outside temperature storage unit 11, starting from the 00 pm 19 obtains the outside air temperature T _m of a last 12 hours.
Here, the similarity calculating unit 12 obtains the image G _m of the last 12 hours, we have obtained outside air temperature T _m of a last 12 hours. However, this is only an example, the similarity calculating unit 12, for example, acquires image G _m for the past 24 hours, may be obtained outside air temperature T _m of a last 24 hours.

The similarity calculation unit 12 has N pixels p _{1, m to} p as evaluation values S _{1 to} NS of _N pixels p _{1, m to p N, m included in the image G m.} The degree of similarity between the temporal transition of each pixel value pd _{1, m to pd N, m in N and m and} the temporal transition of the outside temperature T _m is calculated (step ST1 in FIG. 6).
For example, a temporal transition of the pixel p _n, pixel values pd _{n, m} of the _m, the evaluation value S _n is the degree of similarity between the temporal changes of the outside air temperature T _m is for example, the following formula (1) It is expressed as.

In the indoor layout estimation apparatus 2 shown in FIG. 2, the similarity calculating unit 12, the pixel p _n, as the evaluation value S _n of _m, it calculates the degree of similarity represented by the formula (1). However, well-pixels p _n, as the evaluation value S _n of _m, the pixel values pd _n, and temporal changes in _m, if it is possible to calculate the similarity between the temporal changes of the outside air temperature T _m, the formula ( It is not limited to the degree of similarity represented by 1).
Figure 8 is an explanatory diagram showing an example of calculation results of evaluation value S ₁ to S _N by the similarity calculation unit 12. 8, a region closer to black, indicating a similarity evaluation value S _n is higher regions.
Similarity calculation unit 12 outputs the respective evaluation values S ₁ to S _N of calculated in the window region judgment section 13.

Window region judgment section 13, from the image storage unit 10, among the image _G 1 ~G _M of M where the imaging time t are different from each other, to obtain any one of the images _{G m.} One image G _m of the window region judgment section 13 obtains can be any of the image of the M image G ₁ ~G _M. Window region judgment section 13 may, for example, an imaging time t may acquire the image G _M of the latest time, also the imaging time t acquires image G _M-1 of the old time than the latest time good.
Window region judgment section 13 compares the respective evaluation values _S 1 to S _N calculated by the similarity calculation unit 12, first the threshold Th _1.
Window region judgment section 13, a pixel _{p n} included in the image _{G m obtained,} larger evaluation value _{S n} of _m than the first threshold value Th ₁ (step of FIG. 6 ST2: YES), of _{It is determined that the pixels pn and m} are the pixels in the window area W _m (step ST3 in FIG. 6).
Pixel _{p n,} if the evaluation value _{S n} of the _m first threshold Th ₁ or less (step of FIG. 6 ST2: in the case of NO), the pixel _{p n, m} is a window area _{W m} outside the pixel.

The window area determination unit 13 has determined that all the pixels p _{among the N pixels p 1, m to p N, m included in the acquired image G m} are the pixels in the window area W _{m. It is determined that the region including n and m} is the window region W _m in the image G _m (step ST4 in FIG. 6).
FIG. 9 is an explanatory diagram showing an example of a determination result of the _{window area W m} by the window area determination unit 13. In FIG. 9, the shaded area is the window area W _m . In FIG. 9, the region closer to white is the region where the temperature is higher.
As a determination result of the _{window area W m} , the window area determination unit 13 _{inputs, for example, data indicating coordinates in the image G m in which the window area W m} exists to each of the control unit 3 and the wall area determination unit 17. Output.

Pixel group identifying unit 14a of the standard deviation calculation unit 14, from the image storage unit 10, the imaging time t to obtain the M different image _G 1 ~G _M together.
The pixel group specifying unit 14a _{includes pixels p 1} , 1 to p _{N, 1 , p 1 , 2 to} p N, 2, ..., P ₁ included in each of the _{M images G 1 to} GM. _{, M to} p _{N, M} , each of N pixel groups (1) to (N) including pixels having the same pixel position is specified.
_{Pixels p 1,1 to p N, 1} , p _{1 , 2, to p N, 2} , ..., P _{1, M to p N, M} included in each of the _M images G _{1 to} GM. Among them, for example, M pixels _{pn, 1 to pn, M} are pixels having the same pixel position and are included in the pixel group (n).

偏差算出処理部１４ｂは、算出したそれぞれの評価値ｖ_１〜ｖ_Ｎを前景領域判定部１５に出力する。Deviation calculating unit 14b of the standard deviation calculation unit 14, as shown in the following equation (2), as an evaluation value v _n of the specified pixels by the pixel group specifier 14a (n), the pixel group (n) calculated including M pixels _{p n,} 1 _~p n, pixel values in the _{M pd n,} 1 _~pd n, the standard deviation of _M (step ST5 in FIG. 6).

The deviation calculation processing unit 14b outputs the calculated evaluation values v _{1 to v N} to the foreground region determination unit 15.

Foreground region determining unit 15, from the image storage unit 10, the imaging time t to obtain the M different image _G 1 ~G _M together.
The foreground region determination unit 15 compares the respective evaluation values v _{1 to v N} calculated by the standard deviation calculation unit 14 with the second threshold value Th ₂ .
In the foreground region determination unit 15, if the evaluation value v _n (n = 1, ..., N) is larger than _{the second threshold value Th 2 in the N pixel groups (1) to (N) (} In step ST6 of FIG. 6: YES), it is determined that the pixels _{pn and m} included in the pixel group (n) are the pixels in the foreground region F _m (step ST7 of FIG. 6).
A plurality of pixel groups (1) in a ~ (N), the evaluation value _v if _n is a second threshold value Th ₂ or less (step of FIG. 6 ST6: in the case of NO), the included in the pixel group (n) The pixels _{pn and m} are the pixels in the background area B _m.
Pixel _{p n,} pixel values _{pd n} of _m within the background region _{B m, m,} since the time is also passed, hardly changes, pixel group including a pixel _{p n, m} within the background region _{B m} (n) The evaluation value v _{n of} is likely to be equal to or less than the second threshold value Th _2.
On the other hand, the pixel _{p n,} pixel values _{pd n, m} of the _m in the foreground region _{F m,} in order to change over time, the pixel _{p n} in the foreground region _{F m,} a pixel group including _m (n) The evaluation value v _{n of} is likely to be larger than the second threshold value Th _2.
The pixel values pd _{n, m of the pixels pn, m} in the window area W m also change with the passage of time. However, the change in the pixel values pd _{n, m of the pixels pn, m} in the window region W m is extremely small as compared with the change _{in the pixel values pd n, m} due to the movement of a human being or the like. _{Therefore, the evaluation value v n} of the pixel group (n) including the pixels _{pn and m} in the window region W m is likely to be equal to or less than the second threshold value Th _2.

The foreground region determination unit 15 _{includes pixels p 1} , 1 to p _{N, 1 , p 1 , 2 to} p N, 2, ..., P ₁ included in each of the _{M images G 1 to} GM. _{determination,} M _{~p n,} among the _M, all the pixels _{p n} is determined that the pixels in the foreground area _{F m,} a region including the _m is the foreground region _{F m} in each of the images _{G m} (Step ST8 in FIG. 6).
Foreground region determining unit 15, as the determination result of the foreground region F _m, for example, and outputs the foreground region F _m is present, the data indicating the coordinates within the image G _m in the floor area determination unit 16.

10, among the M captured image _G 1 ~G _M, is an explanatory diagram showing an example of a single image _{G m.}
FIG. 11 is an explanatory diagram showing a background region B _m and a foreground region F _m in a certain image G _{m among the M captured images G 1 to} GM. In FIG. 11, the black region is the region _{determined to be the background region B m} , and the white region is the region determined to be the _{foreground region F m.}
In the example of FIG. 11, the foreground region F _m substantially coincides with the region in which a human exists (hereinafter, referred to as “human region H _m”). In FIG. 11, denoted human region H _m is omitted.
In the indoor layout estimation apparatus 2 shown in FIG. 2, as a foreground region F _m coincides with the human region H _m, floor area determination unit 16 to be described later, based on the foreground area F _m, determine the bed area Y _m do.

The floor area determination unit 16 acquires data indicating the coordinates in the image G _{m in which the foreground area F m} exists from the foreground area determination unit 15.
Floor area determination unit 16, based on the data indicating the acquired coordinate, identifies a plurality of pixels p _{n, m} included in each of the foreground region F _m in the M image G ₁ ~G _M.
Floor area determination unit 16, a plurality of pixels p _n included in each of the foreground region F _m, in _m, in the y direction of the image G _m, floor area pixel pixel present in the position of the lower end Specify as.
If the foreground region F _m coincides with the human region H _{m , the position of the lower end in the foreground region F m} corresponds to the position of the human foot, and the position of the human foot corresponds to the position of the floor. Humans usually move over time. Foreground This human positions at each imaging time t is different M pieces of the captured image G ₁ ~G _M, i.e., the position of the foreground region F ₁ to F _M becomes different from each other, which corresponds to the floor position the position of the lower end of the area F ₁ to F _M also means that the different from each other. Therefore, by imaging time t has overall position of the lower end of the foreground region F ₁ to F _M in M different captured image G ₁ ~G _M to each other, it is possible to estimate the floor area Y _m.

Floor area determination unit 16, a region including all of the floor area pixels included in each of the foreground region F ₁ to F _M is determined to be a floor area Y _m (step ST9 of FIG. 6).
As a determination result of the _{floor area Y m} , the floor area determination unit 16 _{inputs, for example, data indicating coordinates in the image G m in which the floor area Y m} exists to each of the control unit 3 and the wall area determination unit 17. Output.
Here, the floor area determination unit 16 specifies a pixel existing at the lower end position in the vertical direction as a floor area pixel. However, this is only an example, and the floor area determination unit 16 uses a plurality of pixels existing around the lower end pixel as floor area pixels in addition to the pixel existing at the lower end position in the vertical direction. You may try to specify.

Figure 12 is an explanatory view showing the lower end of the pixels in the foreground area _F 1 to F _M in the M image _G 1 ~G _M. In FIG. 12, the white region is a region including the lower end pixels in _{each foreground region F m.}
Image is a bottom of the image pixels are collected in each of the foreground region F _m in the M image G ₁ ~G _M Figure 12 indicates.
In FIG. 12, _{the lower end pixel in the foreground area F 1} , the lower end pixel in the foreground area F ₂ , the lower end pixel in the foreground area F ₃ , the lower end pixel in the foreground area F ₄ , and the foreground area F ₅ the lower end of the pixel, the lower end of the pixels in the foreground area F _M-1, and illustrates the lower end of the pixels in the foreground area F _M.
FIG. 13 is an explanatory diagram showing an example of a determination result of the _{floor area Y m} by the floor area determination unit 16. In FIG. 13, the area provided with the horizontal line is the floor area Y _m .
White areas 12 are shown, the floor area Y _m shown in FIG. 13, for convenience of description, it is somewhat deviated, approximately coincide with each other.

Wall area determining unit 17, and the window area W _m determined by the window region judgment section 13, on the basis of the a floor area Y _m determined by the floor area determination unit 16, a wall area K _m in the target area Judgment (step ST10 in FIG. 6).
Wall area determining unit 17, as the determination result of the wall area K _m, for example, outputs wall area K _m is present, the data indicating the coordinates within the image G _m to the control unit 3.

It will be specifically described below determination process wall area K _m by the wall area determination unit 17.
Wall area determining unit 17, from the image storage unit 10, the M image _G 1 ~G _M of the imaging time t is different, to obtain any one of the images _{G m.} One of the captured image G _m of wall area determination unit 17 obtains can be any of the captured image of the M captured image G ₁ ~G _M. Wall area determining unit 17, for example, an imaging time t may acquire the image G _M of the latest time, also the imaging time t acquires image G _M-1 of the old time than the latest time good.
Further, the wall area determination unit 17 acquires data indicating the coordinates in which the _{floor area Y m exists from the floor area determination unit 16.}

The wall area determination unit 17 identifies a plurality of pixels _{pn, m included in the floor area Y m} based on the data indicating the coordinates in which the _{floor area Y m exists.}
As shown in FIG. 13, the wall area determination unit 17 identifies the uppermost pixel in the y direction of the image G _m among the plurality of pixels _{pn and m included in the floor area Y m.} The upper end of the pixels in the bed area Y _m is a plurality contained in the floor area Y _m pixel p _n, of _m, in the x-coordinate of a plurality of identical pixels is the largest pixel y coordinate .. In FIG. 13, nine pixels at the upper end in the _{floor area Y m are illustrated.}
As shown in FIG. 14, the wall area determination unit 17 determines a line including a line connecting each pixel at the upper end in the _{floor area Y m as a boundary line between the floor surface and the wall surface.}
FIG. 14 is an explanatory diagram showing a boundary line between the floor surface and the wall surface, a floor area Y _m , and a window area W _m.
In FIG. 14, the area with horizontal lines is the floor area Y _m , and the area with diagonal lines is the window area W _m .

The wall area determination unit 17 acquires data indicating the coordinates in which the _{window area W m} exists from the window area determination unit 13.
The wall area determination unit 17 identifies the _{window area W m} based on the data indicating the coordinates in which the _{window area W m exists.
In the image G m} , the wall area determination unit 17 has a wall in the area other than the _{window area W m} , which has a larger y coordinate in the y direction of _{the image G m} than the boundary line between the floor surface and the wall surface. It is determined that the area is K _{m. The region where the y coordinate is larger in the y direction of the image G m} than the boundary line between the floor surface and the wall surface is a region above the boundary line between the floor surface and the wall surface in the figure.
Wall area determining unit 17, as the determination result of the wall area K _m, for example, outputs wall area K _m is present, the data indicating the coordinates within the image G _m to the control unit 3.

Control unit 3, from the image storage unit 10, among the image _G 1 ~G _M of M where the imaging time t are different from each other, to obtain any one of the images _{G m.} One of the captured image G _m the control unit 3 acquires can be any of the captured image of the M captured image G ₁ ~G _M. Control unit 3 may, for example, an imaging time t may acquire the image G _M of the latest time, the imaging time t may acquire the image G _M-1 of the old time than the latest time.
_{The control unit 3 acquires the determination result of the window area W m} from the window area determination unit 13, the determination result of the floor area Y _{m from the floor area determination unit 16, and the wall area K m} from the wall area determination unit 17. Acquire the judgment result.
The control unit 3 adjusts the air in the target region by controlling, for example, the air conditioner 4 based on the acquired determination results.

Hereinafter, an example of air adjustment by the control unit 3 will be specifically described.
For example, it is assumed that the set temperature of the window area W _{m is Temp 1} , the set temperature of the floor area Y _m is Temp ₂ , and the set temperature of the wall area K _{m is Temp 3.} Each of the set temperature Temp ₁ , the set temperature Temp ₂ , and the set temperature Temp ₃ may be stored in the internal memory of the control unit 3, or may be supplied from the outside of the air conditioner shown in FIG. You may.
Control unit 3, based on the determination result of the window area _{W m,} N pixels _p 1 contained in the image _{G m,} m _{~p N,} from among the _m, are included in the window region _{W m} Specify a plurality of pixels _{pn, m} .
Control unit 3, a plurality of pixels p _n contained in the window region W _m, the pixel values pd n of _m, the temperature indicated by _m, for example, calculates the average temperature of the window area W _m.
Control unit 3, the average temperature of the window area _{W m} is to match the set temperature Temp ₁ of the window area _{W m,} controls the air conditioner 4.

Control unit 3, based on the determination result of the floor area _{Y m,} N pieces of pixels included in the image _{G m p 1, m ~p N} , from among the _m, are included in the floor area _{Y m} Specify a plurality of pixels _{pn, m} .
Control unit 3, a plurality of pixels p _n included in the floor area Y _m, the pixel values pd n of _m, the temperature indicated by _m, for example, calculates the average temperature of the floor area Y _m.
Control unit 3, the average temperature of the floor area _{Y m} is to match the set temperature Temp ₂ of the floor area _{Y m,} to control the air conditioner 4.

Based on the determination result of the _{wall region K m} , the control unit 3 includes a plurality of pixels p _{1, m to p N, m included in the image G m} , which are included in the wall region K _m. Pixels _{pn and m of} are specified.
Control unit 3, a plurality of pixels p _n included in the wall area K _m, the pixel values pd n of _m, the temperature indicated by _m, for example, calculates the average temperature of the wall area K _m.
Control unit 3, the average temperature of the wall area _{K m} is to match the set temperature Temp ₃ wall area _{K m,} for controlling the air conditioner 4.

Here, the control unit 3 calculates the average temperature of each of _{the window area W m} , the floor area Y _m, and the wall area K _m. However, this is only an example, and the control unit 3 may calculate, for example, the intermediate temperature of each of _{the window area W m} , the floor area Y _m, and the wall area K _m. In this case, the control unit 3, an intermediate temperature, such as the window area W _m is to match the set temperature, such as the window area W _m, it controls the air conditioner 4.

In the above-described first embodiment, as the evaluation value of each of the plurality of pixels included in the image output from the infrared camera 1 that is capturing the target area, the temporal value of each pixel value in the plurality of pixels is temporal. The evaluation value calculated by the similarity calculation unit 12 is the first among the plurality of pixels included in the image and the similarity calculation unit 12 that calculates the similarity between the transition and the temporal transition of the outside temperature. The indoor layout estimation device 2 is configured to include a window area determination unit 13 for determining that a region including pixels larger than the threshold value of is a region where a window exists in the image. Therefore, the indoor layout estimation device 2 can prevent erroneous detection of the window area even if the outside air temperature changes.

Embodiment 2.
In the indoor layout estimation device 2 shown in FIG. 2, the standard deviation calculation unit 14 has _{pixels p 1} , 1 to p _{N, 1 , p 1, 2} to included in each of the _{M images G 1 to} GM. Among p _{N, 2} , ..., p _{1, M to p N, M , each evaluation value v 1} to a plurality of pixel groups (1) to (N) including pixels having the same pixel position. v _N is calculated.
In the second embodiment, each time the M image G ₁ ~G _M and the imaging time t a new image that is different is stored in the image storage unit 10, a plurality of pixel groups (1) to each of the (N) The indoor layout estimation device 2 for calculating the evaluation values v _{1 to v N will be described.} Here, for convenience of explanation, it is assumed that the new image is GM _{+ 1} .

Hereinafter, the indoor layout estimation device 2 according to the second embodiment will be described.
The configuration of the indoor layout estimation device 2 according to the second embodiment is the same as the configuration of the indoor layout estimation device 2 according to the first embodiment, and the configuration diagram showing the indoor layout estimation device 2 according to the second embodiment is FIG. 2.

Infrared camera 1 outputs the M image _G 1 ~G _M and the imaging time t is different from a new image _{G M + 1} to the indoor layout estimation device 2.
The image storage unit 10 of the indoor layout estimation device 2 stores a new image GM _{+ 1} .
_{When the new image GM + 1} is stored in the image storage unit 10, the pixel group specifying unit 14a of the standard deviation calculation unit 14 acquires the _{new image GM + 1 from the image storage unit 10.}
Pixel group identifying unit 14a, from among the M image _G 1 ~G _M previously acquired, for example, most image capturing time t discards the old image _{G 1.}
The pixel group identifying unit 14a, the image is part of the M image _G 1 ~G _M, i.e., not discarded (M-1) pieces of the image _G 2 ~G _M and a new image _{G M + 1} pixel _p 1, 2 are included in each of the M image _G 2 _{~G M + 1} containing _{~p N, 2, p 1,3 ~p} N, 3, ···, p 1, M + 1 ~p N, _{Among M + 1} , each of N pixel groups (1) to (N) including pixels having the same pixel position is specified.

Here, pixel group specifying unit 14a is, among the previously acquired M pieces of image G ₁ ~G _M, most imaging time t is to discard the old image G _1. However, this is only an example, pixel group specifying unit 14a is most imaging time t may be discarded images other than the old image G _1.
The pixel group specifier 14a is, among the M image G ₁ ~G _M previously acquired, may be discarded two or more images. Pixel group identifying unit 14a, when discard the two images, as discard not in (M-2) including a number and an image and a new image _{G M + 1 (M-1} ) pieces of images, the image G ₃ ~ G _{M + 1} N-number of the contained in each pixel _{p 1,3 ~p N, 3, p} 1,4 ~p N, 4, ···, p 1, M + 1 ~p N, among the _{M + 1,} pixel Each of the N pixel groups (1) to (N) including the pixels having the same position is specified.

In the deviation calculation processing unit 14b of the standard deviation calculation unit 14, each time the pixel group identification unit 14a specifies each of the N pixel groups (1) to (N), each pixel is the same as in the first embodiment. As the evaluation values v _{1 to v N} of the groups (1) to (N), the pixel values in _{the M pixels pn, 1 to pn, M} included in the respective pixel groups (1) to (N). Calculate the standard deviations of pd _{n, 1 to pd n, M.}
The deviation calculation processing unit 14b outputs the calculated evaluation values v _{1 to v N} to the foreground region determination unit 15.
Accordingly, in the indoor layout estimation apparatus 2 according to the second embodiment, each time the M image G ₁ ~G _M and the imaging time t is different from a new image G _{M + 1} is stored in the image storage unit 10, the N each evaluation value _v 1 to v _N of the pixel group (1) ~ (N) is updated.

Foreground region determining unit 15, the standard deviation calculating section 14 each evaluation value _v 1 to v _N each time it is updated, comparing each of the evaluation values _v 1 to v _N and the second threshold value Th _2.
The foreground region determination unit 15 is included in the pixel group (n) if _{the evaluation value v n} (n = 1, ..., N) is _{larger than the second threshold value Th 2 as in the first embodiment.} and it has a pixel _{p n, m} are determined foreground region determining unit 15 that the pixels in the foreground region _{F m} is, n pieces of pixel _{p 1} contained in each of the M image _G 2 _{~G M + 1 , 2 to p N, 2} , p _{1 , 3 to} p N, 3, ..., p _{1, M + 1 to p N, M + 1} , all pixels p determined to be pixels in the foreground region F _m It is determined that the _{region including n and m} is the foreground region F _m in the image G _m.
Foreground region determining unit 15, as the determination result of the foreground region F _m, for example, and outputs the foreground region F _m is present, the data indicating the coordinates within the image G _m in the floor area determination unit 16.

In the second embodiment described above, each time the pixel group specifying unit 14a outputs a new image having a different imaging time from the plurality of images from the infrared camera 1, the pixel group specifying unit 14a is newly added to a part of the images. Among the plurality of pixels included in each of the plurality of images including the same image, each of the plurality of pixel groups including the pixels having the same pixel position is specified, and the deviation calculation processing unit 14b determines the pixel group. Each time each pixel group is specified by the specific unit 14a, the standard of the pixel values in the plurality of pixels included in each pixel group is used as the evaluation value of each pixel group specified by the pixel group identification unit 14a. The indoor layout estimation device 2 was configured to calculate the deviation. Therefore, the indoor layout estimation device 2 can prevent erroneous detection of the window region even if the outside air temperature changes, and can determine the foreground region based on a new image.

Embodiment 3.
In the third embodiment, in the foreground region F _m determined by the foreground region determining unit 15 will be described indoor layout estimation device 2 that humans are provided for determining human region determining unit 18 a region that is present ..

FIG. 15 is a configuration diagram showing the indoor layout estimation device 2 according to the third embodiment.
FIG. 16 is a hardware configuration diagram showing the hardware of the indoor layout estimation device 2 according to the third embodiment.
In FIGS. 15 and 16, the same reference numerals as those in FIGS. 2 and 3 indicate the same or corresponding portions, and detailed description thereof will be omitted.
The indoor layout estimation device 2 includes an image storage unit 10, an outside air temperature storage unit 11, a similarity calculation unit 12, a window area determination unit 13, a standard deviation calculation unit 14, a foreground area determination unit 15, a human area determination unit 18, and a floor area. A determination unit 16 and a wall area determination unit 17 are provided.

In the indoor layout estimation apparatus 2 shown in FIG. 16, the foreground region determining unit 15, as the determination result of the foreground area, the foreground area F _m are present, the human data indicating coordinates in the image G _m area determination section 18 Output to.
The human domain determination unit 18 is realized by, for example, the human domain determination circuit 28 shown in FIG.
_{The human region determination unit 18 determines the human region H m} , which is a region in which a human exists, in _{each foreground region F m} determined by the foreground region determination unit 15.
As a determination result of the _{human region H m} , the human region determination unit 18 _{inputs, for example, data indicating coordinates in the image G m in which the human region H m} exists to each of the control unit 3 and the floor area determination unit 16. Output.

In FIG. 15, an image storage unit 10, an outside air temperature storage unit 11, a similarity calculation unit 12, a window area determination unit 13, a standard deviation calculation unit 14, a foreground area determination unit 15, and a human being, which are components of the indoor layout estimation device 2. It is assumed that each of the area determination unit 18, the floor area determination unit 16, and the wall area determination unit 17 is realized by dedicated hardware as shown in FIG. That is, the indoor layout estimation device 2 includes an image storage circuit 20, an outside air temperature storage circuit 21, a similarity calculation circuit 22, a window area determination circuit 23, a standard deviation calculation circuit 24, a foreground area determination circuit 25, and a human area determination circuit 28. It is assumed that it is realized by the floor area determination circuit 26 and the wall area determination circuit 27.

Here, each of the image storage circuit 20 and the outside temperature storage circuit 21 is, for example, a non-volatile or volatile semiconductor memory such as RAM, ROM, flash memory, EPROM, EEPROM, magnetic disk, flexible disk, optical disk, compact disk. , Mini disc, or DVD.
Further, each of the similarity calculation circuit 22, the window area determination circuit 23, the standard deviation calculation circuit 24, the foreground area determination circuit 25, the human area determination circuit 28, the floor area determination circuit 26, and the wall area determination circuit 27 is, for example, simple. One circuit, a composite circuit, a programmed processor, a parallel programmed processor, an ASIC, an FPGA, or a combination thereof is applicable.

The components of the indoor layout estimation device 2 are not limited to those realized by dedicated hardware, but the indoor layout estimation device 2 is realized by software, firmware, or a combination of software and firmware. There may be.
When the indoor layout estimation device 2 is realized by software, firmware, or the like, each of the image storage unit 10 and the outside air temperature storage unit 11 is configured on the memory 31 of the computer shown in FIG. To make the computer execute the processing procedures of the similarity calculation unit 12, the window area determination unit 13, the standard deviation calculation unit 14, the foreground area determination unit 15, the human area determination unit 18, the floor area determination unit 16, and the wall area determination unit 17. Program is stored in the memory 31. Then, the processor 32 of the computer shown in FIG. 5 executes the program stored in the memory 31.

Further, FIG. 16 shows an example in which each of the components of the indoor layout estimation device 2 is realized by dedicated hardware, and FIG. 5 shows an example in which the indoor layout estimation device 2 is realized by software, firmware, or the like. Shows. However, this is only an example, and some components in the indoor layout estimation device 2 may be realized by dedicated hardware, and the remaining components may be realized by software, firmware, or the like.

Next, the operation of the indoor layout estimation device 2 shown in FIG. 15 will be described. However, since it is the same as the indoor layout estimation device 2 shown in FIG. 2 except for the human area determination unit 18, the operation of the human area determination unit 18 will be mainly described here.
Foreground region determining unit 15, as in the first embodiment, shown when it is determined foreground region F _m, as a determination result of the foreground region F _m, foreground region F _m are present, the coordinates in the image G _m The data is output to the human area determination unit 18.

In the indoor layout estimation apparatus 2 shown in FIG. 2, the floor area determination section 16, a foreground region F _m coincides with the human region H _m, based on the foreground area F _m, to determine the floor area Y _m There is. However, since the foreground region F _m and the human region H _m do not always exactly match, there is a possibility that the floor region Y _m is erroneously detected in the determination of the floor region Y _{m based on the foreground region F m.} be.
_{Upon receiving the determination result of the foreground region F m} from the foreground region determination unit 15, the human region determination unit 18 is a region in which a human exists in the _{foreground region F m} by using a known human body detection method. Determine the human domain H _m.
As a known human body detection method, for example, a method using HOG and SVM can be considered. HOG (Histogram of Oriented Gradients) is a feature quantity that histograms the gradient direction of the brightness of a local region. SVM (Support Vector Machine) is a pattern discriminative model using supervised learning.
As a determination result of the _{human region H m} , the human region determination unit 18 _{inputs, for example, data indicating coordinates in the image G m in which the human region H m} exists to each of the control unit 3 and the floor area determination unit 16. Output.

Control unit 3, from the image storage unit 10, among the image _G 1 ~G _M of M where the imaging time t are different from each other, to obtain any one of the images _{G m.} One of the captured image G _m the control unit 3 acquires can be any of the captured image of the M captured image G ₁ ~G _M. Control unit 3 may, for example, an imaging time t may acquire the image G _M of the latest time, the imaging time t may acquire the image G _M-1 of the old time than the latest time.
_{The control unit 3 acquires the determination result of the window area W m} from the window area determination unit 13, the determination result of the floor area Y _{m from the floor area determination unit 16, and the wall area K m} from the wall area determination unit 17. It acquires a determination result, to obtain a determination result of human region H _m from a human area determining unit 18.

The control unit 3 adjusts the air in the target region by controlling, for example, the air conditioner 4 based on the acquired determination results.
For example, it is assumed that the set temperature of the human region H _{m is Temp 4.} The set temperature Temp ₄ may be stored in the internal memory of the control unit 3 or may be supplied from the outside of the air conditioner shown in FIG.
Control unit 3, based on the determination result of human region _{H m,} N pieces of pixels included in the image _{G m p 1, m ~p N} , from among the _m, are included in the human region _{H m} Specify a plurality of pixels _{pn, m} .
Control unit 3, a plurality of pixels p _n contained in the human region H _m, pixel values pd n of _m, the temperature indicated by _m, for example, calculates the average temperature of the human region H _m.
Control unit 3, the average temperature of the human region _{H m} is to match the set temperature Temp ₄ human region _{H m,} and controls the air conditioner 4.

Here, the control unit 3, which calculates the average temperature of the human region H _m. However, this is only an example, the control unit 3, for example, may be calculated intermediate temperature of the human region H _m. In this case, the control unit 3, an intermediate temperature of the human region _{H m} is to match the set temperature Temp ₄ human region _{H m,} and controls the air conditioner 4.

In the third embodiment, the indoor layout estimation device 2 is configured to include a human area determination unit 18 for determining an area in which a human exists in the foreground area determined by the foreground area determination unit 15. bottom. Therefore, the indoor layout estimation device 2 can prevent erroneous detection of the window region even if the outside air temperature changes, and can also adjust the air in the region where humans are present.

It should be noted that, within the scope of the present invention, any combination of embodiments can be freely combined, any component of each embodiment can be modified, or any component can be omitted in each embodiment. ..

The present invention is suitable for an indoor layout estimation device, an indoor layout estimation method, and an air conditioner for determining an area where a window exists in a target area.

1 Infrared camera, 2 Indoor layout estimation device, 3 Control unit, 4 Air conditioner, 10 Image storage unit, 11 Outside temperature storage unit, 12 Similarity calculation unit, 13 Window area determination unit, 14 Standard deviation calculation unit, 14a Pixel group identification Unit, 14b Deviation calculation processing unit, 15 Foreground area determination unit, 16 Floor area determination unit, 17 Wall area determination unit, 18 Human area determination unit, 20 Image storage circuit, 21 Outside temperature storage circuit, 22 Similarity calculation circuit, 23 Window area determination circuit, 24 standard deviation calculation circuit, 25 foreground area determination circuit, 26 floor area determination circuit, 27 wall area determination circuit, 28 human area determination circuit, 31 memory, 32 processors.

Claims (11)

As the evaluation values of each of the plurality of pixels included in the image output from the infrared camera that is capturing the target area, the temporal transition of each pixel value in the plurality of pixels and the temporal time of the outside air temperature. The similarity calculation unit that calculates the similarity with the transition,
Of the plurality of pixels included in the image, the region including the pixels whose evaluation value calculated by the similarity calculation unit is larger than the first threshold value is the region in the image where the window is present. An indoor layout estimation device equipped with a window area determination unit for determining the existence. Among the plurality of pixels included in each of the plurality of images output from the infrared camera, each pixel group is used as an evaluation value of each of the plurality of pixel groups including the plurality of pixels having the same pixel position. A standard deviation calculation unit that calculates the standard deviation of pixel values in a plurality of pixels included in
Among the plurality of pixel groups, a pixel group in which the evaluation value calculated by the standard deviation calculation unit is larger than the second threshold value is specified, and among the plurality of pixels included in each image, the specified pixel group is specified. The indoor layout estimation device according to claim 1, further comprising a foreground area determination unit for determining that the area including the pixels included in the pixel group is the foreground area in each image. The standard deviation calculation unit
When a plurality of images having different imaging times are output from the infrared camera, among the plurality of pixels included in each of the plurality of images, a plurality of pixel groups including pixels having the same pixel position. Pixel group identification part that identifies each, and
As an evaluation value of each pixel group specified by the pixel group specifying unit, a deviation calculation processing unit for calculating a standard deviation of pixel values in a plurality of pixels included in each pixel group is provided. The indoor layout estimation device according to claim 2, which is characterized. The pixel group specifying unit includes a part of the images and the new image each time a new image having a different imaging time from the plurality of images is output from the infrared camera. Among the plurality of pixels included in each of the plurality of images, each of the plurality of pixel groups including the pixels having the same pixel position is specified.
The deviation calculation processing unit is included in each pixel group as an evaluation value of each pixel group specified by the pixel group specifying unit each time each pixel group is specified by the pixel group specifying unit. The indoor layout estimation device according to claim 3, wherein a standard deviation of pixel values in a plurality of pixels is calculated. Among the plurality of pixels included in each foreground area determined by the foreground area determination unit, the target area exists at the lower end position in the direction corresponding to the vertical direction of the space where the target area exists. A floor area determination unit that identifies pixels as floor area pixels and determines that the area including all of the floor area pixels included in each foreground area is the area where the floor exists in each image. The indoor layout estimation device according to claim 2, wherein the indoor layout estimation device is provided. Based on the area where the window is determined by the window area determination unit and the area where the floor is determined by the floor area determination unit, the area where the wall is present is determined in each image. The indoor layout estimation device according to claim 5, further comprising a wall area determination unit. The indoor layout estimation according to claim 2, wherein a human area determination unit for determining a human area is provided as an area in which a human exists in each of the foreground areas determined by the foreground area determination unit. Device. Among the plurality of pixels included in each human area determined by the human area determination unit, the target area exists at the lower end position in the direction corresponding to the vertical direction of the space where the target area exists. A floor area determination unit that identifies pixels as floor area pixels and determines that the area including all of the floor area pixels included in each human area is the area where the floor exists in each image. The indoor layout estimation device according to claim 7, wherein the indoor layout estimation device is provided. Based on the area where the window is determined by the window area determination unit and the area where the floor is determined by the floor area determination unit, the area where the wall is present is determined in each image. The indoor layout estimation device according to claim 8, further comprising a wall area determination unit. The similarity calculation unit sets the evaluation value of each of the plurality of pixels included in the image output from the infrared camera capturing the target area as the temporal transition of each pixel value in the plurality of pixels. , Calculate the similarity with the temporal transition of the outside temperature,
Among the plurality of pixels included in the image, the area including the pixel whose evaluation value calculated by the similarity calculation unit is larger than the first threshold value is included in the image. An indoor layout estimation method that determines that the area has windows. The indoor layout estimation device according to any one of claims 1 to 9, and the indoor layout estimation device.
An air conditioner including a control unit that adjusts the air in the target area based on the determination result of the area in the indoor layout estimation device.

Images