Human body area extraction system, human body area extraction method, and program

Abstract

To provide a human body area extraction system that can detect a body area for correctly extracting the contour of a human body in a thermal image.SOLUTION: A human body area extraction system 10 comprises: a thermal image acquisition unit 11 that acquires temperature distribution data of a space; a one part detection unit 12 that detects an area of at least one part of a human body in the temperature distribution data; and a body area detection unit 13 that detects a body area including a human body area different from the area of at least one part on the basis of a result of the detection performed by the one part detection unit 12.SELECTED DRAWING: Figure 1

Description

[Technical field]

[0001] The present disclosure relates to a human body area extraction system and the like. [Background technology]

[0002] Patent Document 1 discloses a technique for extracting a human body region in a thermal image. Specifically, the patent document discloses an air conditioner having a first extraction means for extracting a pixel group exceeding a first threshold value in a first temperature distribution acquired from a sensor by a first extraction means, and a second extraction means for extracting, as a human body region, a pixel group exceeding a second threshold value in a second temperature distribution having a higher resolution than the first temperature distribution from the pixel group extracted by the first extraction means. [Prior art documents] [Patent documents]

[0003] [Patent Document 1] International Publication No. 2017 / 029762 Summary of the Invention [Problem to be solved by the invention]

[0004] However, in the technology disclosed in Patent Document 1, when a human body is included in a part of the low-resolution pixels extracted by the first extraction means, most of the pixels are in the background region, and the pixels do not exceed the first threshold and may not be the pixels to be extracted as the human body region by the second extraction means. In this case, the outer edge of the human body region extracted by the second extraction means is missing, making it impossible to correctly extract the outline of the human body.

[0005] Therefore, the present disclosure provides a human body region extraction system capable of detecting a body region for correctly extracting the outline of a human body in a thermal image. [Means for solving the problem]

[0006] The human body region extraction system according to the present disclosure includes a thermal image acquisition unit that acquires temperature distribution data in a space, a one-part detection unit that detects a region of at least one part of the human body in the temperature distribution data, and a body region detection unit that detects a body region including a human body region other than the region of the at least one part based on a detection result of the one-part detection unit.

[0007] The human body region extraction method according to the present disclosure includes a thermal image acquisition step of acquiring temperature distribution data in space, a one-part detection step of detecting a region of at least one part of the human body in the temperature distribution data, and a body region detection step of detecting a body region including a human body region other than the region of the at least one part based on a detection result in the one-part detection step.

[0008] A program according to the present disclosure is a program for causing a computer to execute the above-described human body region extraction method.

[0009] These comprehensive or specific aspects may be realized as a system, a method, an integrated circuit, a computer program, or a recording medium such as a computer-readable CD-ROM, or may be realized as any combination of a system, a method, an integrated circuit, a computer program, and a recording medium. Effect of the Invention

[0010] According to the human body region extraction system and the like of the present disclosure, it is possible to detect a body region in a thermal image in order to correctly extract the outline of a human body. [Brief description of the drawings]

[0011] [Figure 1] FIG. 1 is a diagram illustrating an example of a configuration of a human body region extraction system according to an embodiment. [Diagram 2] FIG. 11 is a diagram illustrating an example of temperature distribution data. [Diagram 3] 5 is a flowchart showing an example of the operation of the human body region extraction system according to the embodiment. [Figure 4]FIG. 13 is a diagram showing an example of a moving heat source in temperature distribution data. [Diagram 5] 10 is a flowchart showing an example of a method for determining a body region in the embodiment. [Figure 6] 11A and 11B are diagrams illustrating an example of a face region and a provisional region in temperature distribution data. [Figure 7] FIG. 13 is a diagram showing an example of a body region in temperature distribution data. [Figure 8] 10 is a flowchart showing an example of a method for resetting a threshold temperature in the embodiment. [Figure 9] 13 is a diagram showing an example of a region determined to be a human body region after a threshold temperature in temperature distribution data is reset; FIG. [Figure 10] FIG. 13 is a diagram showing an example of a human body region after an OR operation in temperature distribution data. [Figure 11] FIG. 11 is a diagram showing an example of a human body region extraction method according to another embodiment. [Figure 12] FIG. 11 is a diagram showing an example of a human body region extraction method according to another embodiment. [Figure 13] FIG. 11 is a diagram showing an example of a human body region extraction method according to another embodiment. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0012] Hereinafter, the embodiments will be described in detail with reference to the drawings as appropriate. However, more detailed explanation than necessary may be omitted. For example, detailed explanation of already well-known matters or duplicate explanation of substantially the same configuration may be omitted. This is to avoid the following explanation becoming unnecessarily redundant and to facilitate understanding by those skilled in the art.

[0013] The inventor provides the accompanying drawings and the following description so that those skilled in the art can fully understand the present disclosure, and does not intend for them to limit the subject matter described in the claims.

[0014] (Embodiment) Hereinafter, a human body region extraction system according to an embodiment will be described with reference to FIGS.

[0015] The human body region extraction system 10 is a system for extracting a human body region in a thermal image, and is mounted on an air conditioner, etc. The human body region extraction system 10 targets a person who is standing upright and extracts the human body region of that person.

[0016] The human body region extraction system 10 is a computer having a processor and a memory. The memory is a ROM (Read Only Memory) and a RAM (Random Access Memory), and can store a program executed by the processor. The human body region extraction system 10 includes a thermal image acquisition unit 11, a one-part detection unit 12, a body region detection unit 13, a threshold setting unit 14, and a human body determination unit 15. The thermal image acquisition unit 11, the one-part detection unit 12, the body region detection unit 13, the threshold setting unit 14, and the human body determination unit 15 are realized by a processor that executes a program stored in the memory. The human body region extraction system 10 may be a device in which its components are provided in one housing, or a system in which its components are distributed and arranged in multiple housings. In addition, the human body region extraction system 10 does not have to be mounted on an air conditioner or the like, and may be realized by a server or the like. In this case, the components of the human body region extraction system 10 may be distributed and arranged in multiple servers.

[0017] The thermal image acquisition unit 11 acquires a thermal image. For example, the thermal image is temperature distribution data of the space acquired by a thermosensor or the like in a space such as a room. For this reason, the thermal image acquisition unit 11 acquires temperature distribution data of the space as a thermal image. For example, a thermosensor generates temperature distribution data of the space by scanning the space. In the temperature distribution data, pixels are arranged in a matrix, and each pixel is associated with a temperature. For example, the row direction in the temperature distribution data corresponds to the horizontal direction, and the column direction corresponds to the vertical direction. In the following, the row direction is also called the horizontal direction, and the column direction is also called the vertical direction. As described above, the human body region extraction system 10 extracts the human body region of a person standing upright, so that in the temperature distribution data, the human body region is basically a region that is long in the vertical direction.

[0018] FIG. 2 is a diagram showing an example of the temperature distribution data.

[0019] In FIG. 2, pixels with low temperatures in the temperature distribution data are colored black and pixels with high temperatures are colored white so that people can visually recognize the temperature distribution data. For example, by identifying pixels with temperatures corresponding to the temperature of the human body, the pixels can be extracted as a human body region. However, the clothing part of the human body region or the outer edge part such as the feet may be easily affected by the ambient temperature and may be missing. In response to this, the human body region extraction system 10 includes a part detection unit 12 and a body region detection unit 13, so that the system can detect a body region for correctly extracting the outline of the human body in a thermal image, and also includes a threshold setting unit 14 and a human body determination unit 15, so that the system can correctly extract the outline of the human body from the detected body region.

[0020] The one-part detection unit 12 detects at least one part of the human body in the temperature distribution data. The at least one part of the human body includes, for example, a face area. The operation of the one-part detection unit 12 will be described in detail later.

[0021] The body region detection unit 13 detects a body region including a human body region based on the temperature distribution data. For example, the body region detection unit 13 detects a body region including at least a human body region other than the one part region based on the detection result of the one part detection unit 12 in the temperature distribution data. The operation of the body region detection unit 13 will be described in detail later.

[0022] The threshold setting unit 14 sets a threshold temperature in the body region. Alternatively, the threshold setting unit 14 sets a threshold temperature gradient for the temperature gradient between adjacent pixels in the body region. The operation of the threshold setting unit 14 will be described in detail later.

[0023] The human body determination unit 15 determines that a region in the body region that is equal to or higher than the threshold temperature is a human body region. Alternatively, the human body determination unit 15 determines that a region in the body region that is sandwiched between pixels whose absolute values ​​of temperature gradients between adjacent pixels exceed the threshold temperature gradient is a human body region. The operation of the human body determination unit 15 will be described in detail later.

[0024] Next, the operation of the human body region extraction system 10 will be described with reference to FIG.

[0025] Fig. 3 is a flowchart showing an example of the operation of the human body region extraction system 10 according to the embodiment. Fig. 3 shows the operation after the thermal image acquisition unit 11 acquires temperature distribution data.

[0026] The human body region extraction system 10 detects the background temperature in the temperature distribution data (step S11). Although the method of detecting the background temperature is not particularly limited, the most frequent temperature in the temperature distribution data can be detected as the background temperature. For example, the human body region extraction system 10 detects the background temperature as 24°C.

[0027] Next, the human body region extraction system 10 sets the temperature range of the human body (step S12). For example, the human body region extraction system 10 sets the temperature range of the human body to 26°C to 40°C. The order of detecting the background temperature (step S11) and setting the temperature range of the human body (step S12) may be reversed. For example, after the temperature range of the human body is set, the temperature that is the most frequent value in the temperature distribution data excluding the temperature range of the human body may be detected as the background temperature.

[0028] Next, the human body region extraction system 10 extracts a moving heat source in the temperature range of the human body (step S13). The method of extracting the moving heat source is not particularly limited, but for example, the method described in International Publication No. 2022 / 176732 can be used to extract the moving heat source as shown in FIG.

[0029] FIG. 4 is a diagram showing an example of a moving heat source in temperature distribution data.

[0030] As shown in Figure 4, areas such as the lower body are affected by clothing and other factors, resulting in temperatures outside the set human body temperature range (for example, temperatures between 24°C and 26°C), and are therefore not extracted as human body areas.

[0031] 3, next, the one part detection unit 12 detects a face area as an area of ​​at least one part of the human body (step S14). Although the method of detecting the face area is not particularly limited, for example, since the shoulder width is approximately twice the size of the face width, the face area can be detected based on the relationship between the shoulder width and the face width.

[0032] Next, the body region detection unit 13 detects the body region (step S15). Here, a method of detecting the body region will be described with reference to FIG.

[0033] Fig. 5 is a flowchart showing an example of a body region determining method in the embodiment. As shown in Fig. 1, body region detection unit 13 has first setting unit 13a and second setting unit 13b in order to detect the body region.

[0034] First, the first setting unit 13a sets a provisional region including a human body region different from the at least one part region based on the detection result of the one part detection unit 12 (step S21). For example, the at least one part region includes a face region, and the human body region different from the at least one part region (face region) is a human body region below the face.

[0035] FIG. 6 is a diagram showing an example of a face region and a provisional region in temperature distribution data.

[0036] For example, the first setting unit 13a sets a provisional region below the face region detected by the one part detection unit 12. For example, as shown in Fig. 6, the first setting unit 13a sets the horizontal width of the provisional region to twice the horizontal width of the face region, sets the position of the provisional region such that the horizontal center of the provisional region and the horizontal center of the face region are in the same column and the lower end of the face region is adjacent to the upper end of the provisional region, and sets the lower end of the provisional region to the lower end of the temperature distribution data.

[0037] Next, the second setting unit 13b determines a body region from the provisional region based on the temperature distribution in the provisional region set by the first setting unit 13a in the temperature distribution data. Since the human body region in the provisional region often has a temperature distribution different from that of the background region, the body region can be determined from the provisional region based on the temperature distribution in the provisional region.

[0038] Specifically, the second setting unit 13b performs the processes from step S22 to step S25 shown in FIG. 5 to determine the body region.

[0039] First, the second setting unit 13b sets n to 1 (step S22). n indicates the number of rows from the top of the provisional region in the temperature distribution data, and the second setting unit 13b first performs the process in step S23 on the first row of the provisional region.

[0040] The second setting unit 13b determines whether the temperature variance (degree of variation) of the nth row in the provisional region is smaller than a threshold value (step S23). The threshold value is not particularly limited, but is set to a variance value expected when a human body region exists in one row, for example. When a human body region does not exist in the nth row, the nth row becomes an entire background region, and the temperature variance value becomes smaller than the threshold value. On the other hand, when a human body region exists in the nth row, the nth row contains a mixture of human body regions and background regions, and the temperature variance value becomes larger than the threshold value.

[0041] If the temperature variance value of the nth row in the provisional region is equal to or greater than the threshold value (No in step S23), the second setting unit 13b increments n by 1 (step S24). That is, the second setting unit 13b performs the process in step S23 for the next row. For example, if the temperature variance value of the first row in the provisional region is equal to or greater than the threshold value, the second setting unit 13b determines whether or not the temperature variance value of the second row in the provisional region is smaller than the threshold value. In this way, the second setting unit 13b repeats the process in step S23 while shifting the rows down one row at a time in the provisional region until the temperature variance value becomes smaller than the threshold value.

[0042] If the variance value of the temperature of the nth row in the provisional region is smaller than the threshold value (Yes in step S23), the second setting unit 13b determines the nth row as the lower end of the body region (step S25). For example, the lower end of the body region can be determined as shown in FIG.

[0043] In this way, the second setting unit 13b determines the lower end of the body region based on the temperature distribution in the provisional region set by the first setting unit 13a in the temperature distribution data, thereby determining the body region. Specifically, the second setting unit 13b acquires the temperature distribution in the provisional region set by the first setting unit 13a in the temperature distribution data for each row, calculates the degree of variation of the acquired temperature distribution for each row, and determines the lower end of the body region from the calculated degree of variation. This makes it possible to determine the region (row) where the tendency of the temperature distribution changes as the boundary between the body region and the background region (for example, the lower end of the body region), and specifically, to determine the row where the tendency of the degree of variation of the temperature distribution changes as the lower end of the body region.

[0044] FIG. 7 is a diagram showing an example of a body region in temperature distribution data.

[0045] As shown in FIG. 6 and FIG. 7, the body region is the same as the provisional region except for the lower end, so by determining the lower end of the body region, the body region can be determined from the provisional region.

[0046] In steps S14 and S15, a relatively large body region including a human body region (e.g., a human body region below the face region) that has a certain relationship with the region of a specific part (e.g., face) of the detected human body and is different from the region of the part (e.g., face region) can be detected. Since such a body region includes the outer edge of the human body region, according to the present disclosure, a body region in which the outer edge of the human body region is not missing in the thermal image, that is, a body region for correctly extracting the outer shape of the human body in the thermal image can be detected. Specifically, a body region for correctly extracting the outer shape of the human body below the face can be detected. By performing the extraction process of the human body region described in step S16 of FIG. 3 using such a body region including a human body region with an outer edge not missing, the outer shape of the human body can be correctly extracted.

[0047] Returning to the explanation in Fig. 3, human body region extraction system 10 extracts a human body region from a detected body region. Specifically, human body determination unit 15 determines an area in the body region that is equal to or higher than a threshold temperature as a human body region, and threshold setting unit 14 resets the threshold temperature according to the determination result of human body determination unit 15. This process is repeated, and human body region extraction system 10 extracts a human body region from a detected body region. Here, a method for resetting the threshold temperature will be explained with reference to Fig. 8.

[0048] FIG. 8 is a flowchart showing an example of a method for resetting the threshold temperature according to the embodiment.

[0049] The human body determination unit 15 determines whether the threshold temperature is background temperature+0.1° C. (step S31). For example, the threshold setting unit 14 resets the threshold temperature to a new threshold temperature that is a predetermined temperature (e.g., 0.1° C.) lower than the current threshold temperature according to the determination result of the human body determination unit 15, and performs the process in step S31 to end the process when the reset threshold temperature becomes substantially the same temperature as the background temperature (e.g., when the threshold temperature=background temperature+0.1° C. (Yes in step S31)).

[0050] When the human body determination unit 15 determines that the threshold temperature is not the background temperature + 0.1°C (No in step S31), it determines whether the human body area in the body area is less than a predetermined ratio in the body area (step S32). The predetermined ratio is not particularly limited, but may be set to, for example, 60%.

[0051] When it is determined that the human body region in the body region is less than a predetermined ratio in the body region (Yes in step S32), threshold setting unit 14 resets the threshold temperature to a temperature 0.1° C. lower (step S33). This increases the area determined as a human body region by human body determination unit 15. Note that 0.1° C. is just an example of the predetermined temperature when setting a temperature lower than the threshold temperature by a predetermined temperature as the new threshold temperature, and is set appropriately.

[0052] Then, the processing is performed again from step S31, and if it is determined in step S32 that the human body area within the body region is equal to or greater than a predetermined percentage of the body region, the processing ends, and the area with a temperature equal to or greater than the threshold temperature set when it is determined that the human body area within the body region is equal to or greater than a predetermined percentage of the body region is extracted as the human body region.

[0053] Fig. 9 is a diagram showing an example of an area determined to be a human body area after resetting the threshold temperature in the temperature distribution data. For example, when the threshold temperature is high, the area determined to be a human body area is small as shown in Fig. 7, but by gradually lowering the threshold temperature, the area determined to be a human body area becomes a predetermined ratio or more as shown in Fig. 9.

[0054] In this way, the threshold setting unit 14 resets the threshold temperature when the area determined to be a human body area by the human body determination unit 15 is less than a predetermined ratio of the body area. If the area determined to be a human body area is small, it is considered that the human body area is missing, so the threshold temperature can be reset. Also, by resetting the threshold temperature to gradually smaller temperatures, the area determined to be a human body area gradually becomes larger, and the threshold temperature can be set so that the human body area is not missing.

[0055] In addition, the lower body is generally cooler than the upper body, and the threshold temperature may be reset by utilizing the characteristic that the area determined to be a human body area extends downward within the body area as the threshold temperature decreases. Specifically, the threshold setting unit 14 may reset the threshold temperature when the number of pixels in the vertical direction of the area determined to be a human body area by the human body determination unit 15 is less than a predetermined number of pixels. The predetermined number of pixels is not particularly limited, but may be set according to the number of pixels in the vertical direction of the body area, for example. In this way, when the number of pixels in the vertical direction of the area determined to be a human body area is small, it is considered that the human body area is missing, and therefore the threshold temperature can be reset.

[0056] In addition, the threshold setting unit 14 may reset the threshold temperature if the amount of change in the area determined as a human body area by the human body determination unit 15 before and after resetting the threshold temperature is less than a predetermined ratio to the area determined as a human body area by the human body determination unit 15 before resetting the threshold temperature. In this case, for example, the minimum temperature in the body area may be set as the initial value of the threshold temperature, and the threshold temperature may be gradually increased until the amount of change (amount of decrease) becomes equal to or greater than a predetermined ratio. In other words, the threshold setting unit 14 may reset the threshold temperature to a temperature that is a predetermined temperature higher than the threshold temperature according to the determination result of the human body determination unit 15. When the threshold temperature is the initial value (i.e., the threshold temperature is the minimum temperature in the body area), all pixels in the body area are determined to be human body areas, but as the threshold temperature increases, there is a time when background temperatures that are relatively frequent in the body area are excluded altogether from pixels determined to be human body areas. At this timing, many pixels determined to be human body regions are removed, the amount of change (amount of decrease) becomes equal to or exceeds a predetermined ratio, and the resetting of the threshold temperature is completed. In this way, if the amount of change in the region determined to be a human body region before and after the resetting of the threshold temperature is small, it is considered that the region determined to be a human body region within the body region still contains background pixels, so the resetting of the threshold temperature can be further performed.

[0057] In step S16, the threshold temperature can be reset so that the human body region is not missing within a relatively large body region that includes pixels of the human body region whose outer edge is not missing, and the human body region is determined based on the threshold temperature reset so that the human body region is not missing, thereby allowing the outline of the human body to be correctly extracted from the thermal image. Correct extraction of the outline of the human body allows for improved accuracy in detecting the surface temperature of the human body and improved accuracy in estimating the thermal sensation.

[0058] The threshold setting unit 14 may set a threshold temperature gradient for the temperature gradient with adjacent pixels in the body region, and the human body determination unit 15 may determine that a region sandwiched between pixels whose absolute values ​​of the temperature gradient with adjacent pixels exceed the threshold temperature gradient in the body region is a human body region, and the threshold setting unit 14 may reset the threshold temperature gradient according to the determination result of the human body determination unit 15. Since there is a temperature difference between the human body region and the background region, a pixel having a temperature difference with adjacent pixels can be estimated as the boundary between the human body region and the background region, and further, the region sandwiched between the pixels can be estimated as a human body region. In other words, even when the threshold temperature gradient is used, the human body region can be determined. Therefore, the threshold temperature in the above description can be replaced with the threshold temperature gradient.

[0059] Returning to the explanation of FIG. 3, the human body region extraction system 10 performs an OR operation between the heat source extracted in step S13 and the human body region extracted in step S16 (step S17).

[0060] FIG. 10 is a diagram showing an example of a human body region after an OR operation in temperature distribution data.

[0061] By combining the heat source shown in FIG. 4 with the human body region shown in FIG. 9, it is possible to extract a human body region that includes at least a part of the human body (e.g., a face region), as shown in FIG. 10.

[0062] (Other embodiments) As described above, the embodiments have been described as examples of the technology disclosed in this application. However, the technology in this disclosure is not limited to these, and can be applied to embodiments in which modifications, substitutions, additions, omissions, etc. are appropriately made. In addition, it is also possible to combine the components described in the above embodiments to create new embodiments.

[0063] For example, in the above embodiment, an example has been described in which the at least one region includes a face region, but the at least one region does not have to include a face region. For example, the at least one region does not have to include a face region as long as it includes a region of a part having a certain relationship with another human body region.

[0064] For example, in the above embodiment, an example has been described in which human body region extraction system 10 includes threshold setting unit 14 and human body determination unit 15, but human body region extraction system 10 does not have to include threshold setting unit 14 and human body determination unit 15. In this case, human body region extraction system 10 becomes a system for detecting a body region for correctly extracting the outer shape of a human body.

[0065] For example, in the above embodiment, an example has been described in which human body region extraction system 10 includes one-part detection unit 12, but human body region extraction system 10 does not need to include one-part detection unit 12, and there is no particular limitation on the method by which body region detection unit 13 detects the body region. In this case, human body region extraction system 10 becomes a system for correctly extracting the external shape of the human body.

[0066] For example, the present disclosure can be realized not only as the human body region extraction system 10, but also as a human body region extraction method including steps (processing) performed by the components that make up the human body region extraction system 10.

[0067] 11 to 13 are diagrams showing an example of a human body region extraction method according to another embodiment.

[0068] For example, as shown in FIG. 11, the human body region extraction method includes a thermal image acquisition step (step S101) for acquiring temperature distribution data in space, a one-part detection step (step S102) for detecting at least one part of a human body region in the temperature distribution data, and a body region detection step (step S103) for detecting a body region including a human body region other than the at least one part of the human body region based on a detection result in the one-part detection step.

[0069] For example, as shown in FIG. 12, the human body region extraction method includes a thermal image acquisition step (step S201) for acquiring temperature distribution data in space, a body region detection step (step S202) for detecting a body region including a human body region based on the temperature distribution data, a threshold setting step (step S203) for setting a threshold temperature within the body region, and a human body determination step (step S204) for determining that a region within the body region having a temperature equal to or higher than the threshold temperature is a human body region. In the threshold setting step (step S203), the threshold temperature is reset depending on the determination result in the human body determination step.

[0070] For example, as shown in FIG. 13, the human body region extraction method includes a thermal image acquisition step (step S301) for acquiring temperature distribution data in space, a body region detection step (step S302) for detecting a body region including a human body region based on the temperature distribution data, a threshold setting step (step S303) for setting a threshold temperature gradient for a temperature gradient with adjacent pixels within the body region, and a human body determination step (step S304) for determining, as a human body region, a region sandwiched between pixels whose absolute values ​​of temperature gradients with adjacent pixels exceed the threshold temperature gradient, and in the threshold setting step (step S303), the threshold temperature gradient is reset depending on the determination result in the human body determination step.

[0071] For example, the present disclosure can be realized as a program for causing a computer (processor) to execute steps included in the human body region extraction method. Furthermore, the present disclosure can be realized as a non-transitory computer-readable recording medium, such as a CD-ROM, on which the program is recorded.

[0072] For example, when the present disclosure is realized by a program (software), each step is executed by executing the program using hardware resources such as a computer's CPU, memory, input / output circuit, etc. In other words, each step is executed by the CPU acquiring data from the memory or input / output circuit, etc., performing calculations, and outputting the calculation results to the memory or input / output circuit, etc.

[0073] In the above embodiment, each component included in the human body region extraction system 10 may be configured with dedicated hardware, or may be realized by executing a software program suitable for each component. Each component may be realized by a program execution unit such as a CPU or a processor reading and executing a software program recorded on a recording medium such as a hard disk or a semiconductor memory.

[0074] A part or all of the functions of the human body region extraction system 10 according to the above embodiment are typically realized as an LSI, which is an integrated circuit. These may be individually integrated into one chip, or may be integrated into one chip that includes some or all of the functions. The integrated circuit is not limited to an LSI, and may be realized by a dedicated circuit or a general-purpose processor. An FPGA (Field Programmable Gate Array) that can be programmed after the LSI is manufactured, or a reconfigurable processor that can reconfigure the connections and settings of circuit cells inside the LSI may be used.

[0075] Furthermore, if an integrated circuit technology that can replace LSI emerges due to advances in semiconductor technology or a different derivative technology, the components included in the human body region extraction system 10 may naturally be integrated into integrated circuits using that technology.

[0076] In addition, the present disclosure also includes forms obtained by applying various modifications to the embodiments that a person skilled in the art may conceive, and forms realized by arbitrarily combining the components and functions of each embodiment within the scope that does not deviate from the spirit of the present disclosure.

[0077] (Additional Note) The above description of the embodiments discloses the following techniques.

[0078] (Technology 1) A human body region extraction system comprising: a thermal image acquisition unit that acquires temperature distribution data in a space; a one-part detection unit that detects a region of at least one part of a human body in the temperature distribution data; and a body region detection unit that detects a body region including a human body region other than the region of the at least one part based on a detection result of the one-part detection unit.

[0079] According to this, it is possible to detect a relatively large body region including a human body region (e.g., a human body region below the face region) that has a certain relationship with the region of a specific part (e.g., face) of the detected human body and is different from the region of the part (e.g., face region).Since such a body region includes pixels of the outer edge of the human body region, according to the present disclosure, it is possible to detect a body region in which the outer edge of the human body region is not missing in the thermal image, that is, a body region for correctly extracting the outline of the human body in the thermal image.

[0080] (Technology 2) The human body region extraction system according to Technology 1, wherein the body region detection unit includes a first setting unit that sets a provisional region including a human body region different from the region of the at least one part based on a detection result of the one part detection unit, and a second setting unit that determines the body region from within the provisional region based on a temperature distribution in the temperature distribution data within the provisional region set by the first setting unit.

[0081] According to this, since the human body region within the provisional region often has a different temperature distribution from the background region, the body region can be determined by determining the boundary between the human body region and the background region within the provisional region based on the temperature distribution within the provisional region.

[0082] (Technology 3) The human body region extraction system according to Technology 2, wherein the second setting unit determines the body region by determining a lower end of the body region based on a temperature distribution in the provisional region set by the first setting unit in the temperature distribution data.

[0083] This makes it possible to determine the region where the temperature distribution trend changes to be the boundary between the body region and the background region (for example, the lower end of the body region).

[0084] (Technology 4) The human body region extraction system according to Technology 3, wherein the second setting unit acquires a temperature distribution within the provisional region set by the first setting unit for each row in the temperature distribution data, calculates a degree of variation in the acquired temperature distribution for each row, and determines a lower end of the body region from the calculated degree of variation.

[0085] According to this, the row where the tendency of the degree of variation in temperature distribution changes can be determined as the lower end of the body region.

[0086] (Technology 5) The human body region extraction system according to any one of Technologies 1 to 4, wherein the at least one region includes a face region, and the other human body region than the at least one region is a human body region below the face region.

[0087] This makes it possible to detect the body region for correctly extracting the outline of the human body below the face.

[0088] (Technology 6) The human body region extraction system according to any one of Technologies 1 to 5, further comprising a threshold setting unit that sets a threshold temperature within the body region, and a human body determination unit that determines an area within the body region that has a temperature equal to or higher than the threshold temperature as a human body region, wherein the threshold setting unit resets the threshold temperature depending on a determination result of the human body determination unit.

[0089] According to this, the threshold temperature can be reset so that the human body region is not missing, and the human body region can be determined based on the threshold temperature reset so that the human body region is not missing, so that the outline of the human body can be correctly extracted from the thermal image. By correctly extracting the outline of the human body, it is possible to improve the detection accuracy of the surface temperature of the human body and the estimation accuracy of the thermal sensation.

[0090] (Technology 7) The human body region extraction system described in Technology 6, wherein the threshold setting unit resets the threshold temperature when the area determined to be a human body region by the human body determination unit is less than a predetermined proportion of the body region.

[0091] According to this, if there are few areas determined to be human body areas, it is considered that there are missing human body areas, and therefore the threshold temperature can be reset.

[0092] (Technology 8) The human body region extraction system described in Technology 6, wherein the threshold setting unit resets the threshold temperature when the number of vertical pixels in the region determined to be a human body region by the human body determination unit is less than a predetermined number of pixels.

[0093] According to this, when the number of pixels in the vertical direction in an area determined to be a human body area is small, it is considered that the human body area is missing, and therefore the threshold temperature can be reset.

[0094] (Technology 9) The human body region extraction system described in Technology 6, wherein the threshold setting unit resets the threshold temperature when the amount of change in the region determined to be a human body region by the human body determination unit before and after resetting the threshold temperature is less than a predetermined ratio compared to the region determined to be a human body region by the human body determination unit before resetting the threshold temperature.

[0095] According to this, if the amount of change in the area determined to be a human body area before and after the threshold temperature is reset is small, it is considered that the human body area is still missing or that the human body area contains background pixels, so the threshold temperature can be further reset.

[0096] (Technology 10) A human body region extraction system according to any one of Technologies 6 to 9, wherein the threshold setting unit resets the threshold temperature to a temperature that is a predetermined temperature lower than the threshold temperature or a predetermined temperature higher than the threshold temperature depending on the judgment result of the human body judgment unit.

[0097] According to this, by resetting the threshold temperature to a gradually lower or higher temperature, the area determined to be a human body area gradually becomes larger, and the threshold temperature can be set so that no human body area is missing.

[0098] (Technology 11) The human body region extraction system according to any one of Technologies 1 to 5, further comprising a threshold setting unit that sets a threshold temperature gradient for a temperature gradient with adjacent pixels in the body region, and a human body determination unit that determines, as a human body region, a region sandwiched between pixels whose absolute values ​​of temperature gradients with adjacent pixels exceed the threshold temperature gradient, wherein the threshold setting unit resets the threshold temperature gradient depending on a determination result of the human body determination unit.

[0099] According to this, the threshold temperature gradient can be reset so that the human body region is not missing, and the human body region can be determined by the threshold temperature gradient reset so that the human body region is not missing, so that the outline of the human body can be correctly extracted from the thermal image. By correctly extracting the outline of the human body, it is possible to improve the detection accuracy of the surface temperature of the human body and the estimation accuracy of the thermal sensation.

[0100] (Technology 12) The human body region extraction system described in Technology 11, wherein the threshold setting unit resets the threshold temperature gradient when the area determined to be a human body region by the human body determination unit is less than a predetermined proportion of the body region.

[0101] According to this, if the number of areas determined to be human body areas is small, it is considered that the human body areas are missing, and therefore the threshold temperature gradient can be reset.

[0102] (Technology 13) The human body region extraction system described in Technology 11, wherein the threshold setting unit resets the threshold temperature gradient when the number of vertical pixels in the region determined to be a human body region by the human body determination unit is less than a predetermined number of pixels.

[0103] According to this, when the number of pixels in the vertical direction in an area determined to be a human body area is small, it is considered that the human body area is missing, and therefore the threshold temperature gradient can be reset.

[0104] (Technology 14) The threshold setting unit resets the threshold temperature gradient when the amount of change in the area determined to be a human body area by the human body determination unit before and after resetting the threshold temperature gradient is less than a predetermined ratio compared to the area determined to be a human body area by the human body determination unit before resetting the threshold temperature gradient. This is the human body area extraction system described in Technology 11.

[0105] According to this, if the amount of change in the area determined to be a human body area before and after the resetting of the threshold temperature gradient is small, it is considered that the human body area is still missing or that the human body area contains background pixels, so the threshold temperature gradient can be further reset.

[0106] (Technology 15) A human body region extraction system according to any one of Technologies 11 to 14, wherein the threshold setting unit resets the threshold temperature gradient to a temperature gradient that is a predetermined temperature smaller than the threshold temperature gradient, or a temperature gradient that is a predetermined temperature larger than the threshold temperature gradient, depending on the judgment result of the human body judgment unit.

[0107] According to this, by gradually resetting the threshold temperature gradient to a smaller or larger temperature gradient, the area determined to be a human body area gradually becomes larger, and the threshold temperature gradient can be set so that no human body area is missing.

[0108] (Technology 16) A human body region extraction method including: a thermal image acquisition step for acquiring temperature distribution data in a space; a one-part detection step for detecting a region of at least one part of a human body in the temperature distribution data; and a body region detection step for detecting a body region including a human body region other than the region of the at least one part based on a detection result in the one-part detection step.

[0109] This provides a human body region extraction method capable of detecting a body region for correctly extracting the outline of a human body in a thermal image.

[0110] (Technology 17) A program for causing a computer to execute the human body region extraction method described in Technology 16.

[0111] This makes it possible to provide a program capable of detecting body regions in thermal images in order to correctly extract the outline of a human body. [Industrial Applicability]

[0112] The present disclosure can be applied to systems that detect people using thermal images. [Explanation of symbols]

[0113] 10. Human body region extraction system 11 Thermal image acquisition unit 12 One-part detection unit 13 Body region detection unit 13a First Setting Section 13b Second Setting Section 14 Threshold setting section 15 Human Body Identification Department

Claims

1. A thermal image acquisition unit for acquiring temperature distribution data in a space; a part detection unit that detects an area of ​​at least one part of a human body in the temperature distribution data; a body region detection unit that detects a body region including a human body region other than the at least one part region based on a detection result of the one part detection unit, Human body region extraction system.

2. The body region detection unit is a first setting unit that sets a provisional region including a human body region different from the region of the at least one part based on a detection result of the one part detection unit; a second setting unit that determines the body region from within the provisional region based on a temperature distribution in the temperature distribution data within the provisional region that is set by the first setting unit, The human body region extraction system according to claim 1 .

3. the second setting unit determines the body region by determining a lower end of the body region based on a temperature distribution in the provisional region set by the first setting unit in the temperature distribution data. The human body region extraction system according to claim 2 .

4. The second setting unit acquires a temperature distribution in the provisional region set by the first setting unit for each row in the temperature distribution data, calculates a degree of variation in the acquired temperature distribution for each row, and determines a lower end of the body region from the calculated degree of variation. The human body region extraction system according to claim 3 .

5. The at least one region includes a face region, The human body region other than the at least one region is a human body region below a face region. The human body region extraction system according to claim 1 .

6. moreover, a threshold setting unit that sets a threshold temperature within the body region; a human body determination unit that determines an area in the body area that has a temperature equal to or higher than the threshold temperature as a human body area, The threshold setting unit resets the threshold temperature in accordance with a determination result of the human body determination unit. The human body region extraction system according to any one of claims 1 to 5.

7. the threshold setting unit resets the threshold temperature when the area determined to be a human body area by the human body determination unit is less than a predetermined ratio to the body area. The human body region extraction system according to claim 6 .

8. the threshold setting unit resets the threshold temperature when the number of pixels in the vertical direction of the region determined to be a human body region by the human body determination unit is less than a predetermined number of pixels. The human body region extraction system according to claim 6 .

9. the threshold setting unit resets the threshold temperature when a change in an area determined to be a human body area by the human body determination unit before and after resetting the threshold temperature is less than a predetermined ratio to an area determined to be a human body area by the human body determination unit before resetting the threshold temperature. The human body region extraction system according to claim 6 .

10. the threshold setting unit resets the threshold temperature to a temperature that is a predetermined temperature lower than the threshold temperature or a predetermined temperature higher than the threshold temperature in accordance with a determination result of the human body determination unit. The human body region extraction system according to claim 6 .

11. moreover, a threshold setting unit that sets a threshold temperature gradient for a temperature gradient between adjacent pixels in the body region; a human body determination unit that determines, within the body region, a region sandwiched between pixels whose absolute values ​​of temperature gradients with adjacent pixels exceed the threshold temperature gradient as a human body region; The threshold setting unit resets the threshold temperature gradient in accordance with a determination result of the human body determination unit. The human body region extraction system according to any one of claims 1 to 5.

12. the threshold setting unit resets the threshold temperature gradient when the area determined to be a human body area by the human body determination unit is less than a predetermined ratio to the body area. The human body region extraction system according to claim 11.

13. the threshold setting unit resets the threshold temperature gradient when the number of pixels in the vertical direction of the region determined to be a human body region by the human body determination unit is less than a predetermined number of pixels. The human body region extraction system according to claim 11.

14. the threshold setting unit resets the threshold temperature gradient when an amount of change in the area determined to be a human body area by the human body determination unit before and after resetting the threshold temperature gradient is less than a predetermined ratio to the area determined to be a human body area by the human body determination unit before resetting the threshold temperature gradient. The human body region extraction system according to claim 11.

15. the threshold setting unit resets the threshold temperature gradient to a temperature gradient that is smaller by a predetermined temperature than the threshold temperature gradient or a temperature gradient that is larger by a predetermined temperature than the threshold temperature gradient, depending on the determination result of the human body determination unit. The human body region extraction system according to claim 11.

16. A thermal image acquisition step for acquiring temperature distribution data in a space; a part detection step of detecting an area of ​​at least one part of a human body in the temperature distribution data; and a body region detection step of detecting a body region including a human body region other than the at least one part region based on a detection result in the one part detection step. Human body region extraction method.

17. A program for causing a computer to execute the human body region extraction method according to claim 16.

Images