Passing person counting device

Abstract

To provide a passing people number measuring device that is not easily affected by an external factor.SOLUTION: A passing people number measuring device 10A measuring the number of people passing through a detection area A comprises: a thermal image creation unit 11 that creates a thermal image related to the detection area A every time a time Δt elapses; a difference image creation unit 21 that creates a first difference image that is a difference image between a thermal image created at a time t and a thermal image created at a time t+Δt; a first noise removal processing unit 22 that executes first noise removal processing on the first difference image; and a measuring unit 23 that measures the number of people passing through the detection area on the basis of a second difference image that is a difference image after the first noise removal processing. The first noise removal processing is processing of measuring, of pixels constituting the first difference image, the number of pixels indicating temperature rise and the number of pixels indicating temperature drop, and if there is a cell in which the two measured numbers or a ratio of the numbers satisfies a predetermined determination condition, changing all pixels constituting the cell to pixels indicating no temperature rise and no temperature drop.SELECTED DRAWING: Figure 1

Description

Technical Field

[0001] The present invention relates to a passing person counting device that measures the number of people passing through a detection area provided near entrances and exits of various facilities and various vehicles.

Background Art

[0002] Conventionally, an entrance and exit monitoring device described in Patent Document 1 has been known. This device includes a pair of thermal sensors (outdoor side thermopile sensor and indoor side thermopile sensor) provided above an entrance and exit, an amplifier that amplifies the difference between the output signals of each sensor, and a CPU processing unit that discriminates human movement based on the output signal of the amplifier.

[0003] According to this device, not only entry and exit of a person, but also the fact that a person trying to enter made a U-turn without entering, and the fact that a person trying to exit made a U-turn without exiting can be discriminated. However, this device cannot accurately identify the number of people when a plurality of people enter or exit simultaneously. The same applies when entry and exit occur simultaneously.

[0004] In order to solve this problem, a type of device has also been proposed that includes a thermal image sensor that generates a thermal image of a detection area provided near an entrance and exit at regular intervals, and a discrimination unit that discriminates human movement based on the generated thermal image.

[0005] According to this device, the passing direction and the number of passing people of people passing through the detection area can be discriminated. For example, according to this device, it can be discriminated that two people enter from an entrance and exit and three people go out from the entrance and exit. However, this device cannot accurately discriminate human movement when the temperature of the detection area changes due to external factors such as natural wind, artificial wind such as air conditioning, and sunlight.

Prior Art Documents

Patent Documents

[0006] [Patent Document 1] Japanese Patent Application Publication No. 9-113355 [Overview of the project] [Problems that the invention aims to solve]

[0007] This invention has been made in view of the above circumstances, and aims to provide a pass-by counting device that is less susceptible to external factors. [Means for solving the problem]

[0008] To solve the above problems, the first person-passing-by-a-person counting device according to the present invention is a device for measuring the number of people who have passed through a predetermined detection area, and comprises: a thermal image generation unit that generates a thermal image relating to the detection area every time Δt has elapsed; a difference image generation unit that generates a first difference image which is a difference image between the thermal image generated at time t and the thermal image generated at time t+Δt; a first noise reduction processing unit that applies a first noise reduction processing to the first difference image; and a measurement unit that measures the number of people who have passed through the detection area based on a second difference image which is a difference image after the first noise reduction processing, wherein the first noise reduction processing measures the number of pixels indicating temperature rise and the number of pixels indicating temperature fall in each of a plurality of cells constituting the first difference image, and if there is a cell in which the two measured numbers or their ratio satisfy a predetermined determination condition, all pixels constituting that cell are changed to pixels that do not indicate temperature rise or temperature fall.

[0009] Furthermore, in order to solve the above problems, the second person-passing measurement device according to the present invention is a device for measuring the number of people who have passed through a predetermined detection area, and comprises: a thermal image generation unit that generates a thermal image relating to the detection area each time time Δt has elapsed; a difference image generation unit that generates a first difference image which is the difference image between the thermal image generated at time t and the thermal image generated at time t+Δt; a determination processing unit that performs a determination process to determine whether or not a first noise reduction process is necessary based on the first difference image; a first noise reduction processing unit that performs a first noise reduction process on the first difference image if it is determined that a first noise reduction process is necessary; a second noise reduction processing unit that performs a second noise reduction process on the first difference image if it is determined that a first noise reduction process is unnecessary; and a second difference image which is the difference image after the first noise reduction process, or a third difference image which is the difference image after the second noise reduction process. The system comprises a measurement unit that measures the number of people who have passed through a detection area, and the determination process measures the number of pixels indicating temperature rise and the number of pixels indicating temperature fall in the entire first difference image, and if the two measured numbers or their ratio satisfy predetermined determination conditions, it is determined that a first noise reduction process is necessary, and if not, it is determined that a first noise reduction process is unnecessary. The first noise reduction process measures the number of pixels indicating temperature rise and the number of pixels indicating temperature fall in each of the multiple cells constituting the first difference image, and if there is a cell in which the two measured numbers or their ratio satisfy predetermined determination conditions, it is changed all the pixels constituting that cell to pixels that do not indicate temperature rise or temperature fall. The second noise reduction process is a process that changes all the pixels constituting the first difference image to pixels that do not indicate temperature rise or temperature fall.

[0010] The first noise reduction process described above may, for example, involve measuring the number of pixels indicating temperature increase and the number of pixels indicating temperature decrease in each of the multiple cells constituting the first difference image, and if there is a cell in which either of the two measured numbers is smaller than a predetermined lower limit, then changing all the pixels constituting that cell to pixels that do not indicate temperature increase or decrease.

[0011] Alternatively, the first noise reduction process described above may involve, for example, measuring the number of pixels indicating temperature increase and the number of pixels indicating temperature decrease in each of the multiple cells constituting the first difference image, and if there is a cell in which the ratio of the larger of the two measured numbers to the smaller one is greater than a predetermined maximum ratio, changing all the pixels constituting that cell to pixels that do not indicate temperature increase or decrease.

[0012] Furthermore, the above determination process may also involve, for example, measuring the number of pixels indicating temperature increase and the number of pixels indicating temperature decrease in the entire first difference image, and determining that the first noise reduction process is unnecessary if either of the two measured numbers is smaller than a predetermined lower limit.

[0013] Alternatively, the above determination process may be, for example, a process that measures the number of pixels indicating temperature increase and the number of pixels indicating temperature decrease in the entire first difference image, and determines that the first noise reduction process is unnecessary if the ratio of the larger of the two measured numbers to the smaller one is greater than a predetermined maximum ratio. [Effects of the Invention]

[0014] According to the present invention, it is possible to provide a pass-by counting device that is less susceptible to external factors. [Brief explanation of the drawing]

[0015] [Figure 1] This is a block diagram of a pass-by counting device according to the first embodiment of the present invention. [Figure 2] This figure shows the first difference image generated in the passerby counting device according to the first embodiment. [Figure 3] This figure shows the first noise reduction process in the pass-by person measuring device according to the first embodiment. [Figure 4] This is an operation flowchart of the pedestrian counting device according to the first embodiment. [Figure 5] This is a block diagram of a pass-by counting device according to a second embodiment of the present invention. [Figure 6]It is a diagram showing the first noise removal process and the second noise removal process in the passing person counting device according to the second embodiment. [Figure 7] It is a flowchart of the operation of the passing person counting device according to the second embodiment.

Embodiments for Carrying Out the Invention

[0016] Hereinafter, embodiments of the passing person counting device according to the present invention will be described with reference to the accompanying drawings.

[0017] [First Embodiment] FIG. 1 shows a passing person counting device 10A according to the first embodiment of the present invention. The passing person counting device 10A is for measuring the number of pedestrians W1 passing in the M1 direction and the number of pedestrians W2 passing in the M2 direction in a detection area A provided on the floor 200 near the entrance and exit of a commercial facility. As shown in the figure, it includes a thermal image generation unit 11, a storage unit 12, a control unit 20A, a setting unit 13, an external interface unit 14, an audio output unit 15, a light emitting unit 16, and a housing for accommodating these. The passing person counting device 10A is attached to the ceiling 100 so that the thermal image generation unit 11 faces the detection area A.

[0018] In this embodiment, the mounting height H is 4 m. Also, the detection area A in this embodiment is a rectangle with a side length of 4 m orthogonal to the assumed passing directions M1 and M2 and a side length of 3 m parallel to the passing directions M1 and M2. When the mounting height H changes, the size of the detection area A changes accordingly.

[0019] The thermal image generation unit 11 is composed of a two-dimensional thermal image sensor. The thermal image generation unit 11 detects far-infrared rays radiated from the detection area A (floor 200) or the pedestrians W1 and W2 in the area A, and generates a thermal image of 60 pixels × 45 pixels (more precisely, data related to the thermal image) corresponding to the detection result. The generation of the thermal image is performed every 1 / 10 s. The generated thermal image is sequentially sent to the storage unit 12.

[0020] The storage unit 12 is composed of a volatile or non-volatile memory. The storage unit 12 stores at least the latest and the one before the latest of the thermal images sent from the thermal image generation unit 11 every 1 / 10 s. In other words, the storage unit 12 stores the last generated thermal image (hereinafter referred to as the "second thermal image") and the thermal image generated 1 / 10 s before that (hereinafter referred to as the "first thermal image").

[0021] The control unit 20A is composed of a microprocessor (MPU) and a computer program executed on the processor. In this embodiment, by executing the computer program, a difference image generation unit 21, a first noise removal processing unit 22, and a measurement unit 23 are formed in the control unit 20A.

[0022] Based on the first thermal image and the second thermal image stored in the storage unit 12, the difference image generation unit 21 generates a first difference image P1 of 60 pixels × 45 pixels as shown in FIG. 2 (more precisely, data related to the first difference image P1). For example, the difference image generation unit 21 obtains the temperature change to be indicated by the pixel at the coordinate (0, 0) of the first difference image P1 by subtracting the temperature indicated by the pixel at the coordinate (0, 0) of the first thermal image from the temperature indicated by the pixel at the coordinate (0, 0) of the second thermal image. It can be said that the first difference image P1 indicates what temperature changes occurred during 1 / 10 s.

[0023] The first noise reduction processing unit 22 generates a second difference image P2 (more precisely, data relating to the second difference image P2) by applying a first noise reduction process to the first difference image P1 generated by the difference image generation unit 21. Here, the first noise reduction process measures the number of pixels indicating temperature increase and the number of pixels indicating temperature decrease in each of the 25 cells C00, C10, C20, ..., C24, C34, C44 (see Figure 2) that make up the first difference image P1. If there is a cell in which either of the two measured numbers is smaller than a predetermined lower limit, all pixels constituting that cell are changed to pixels that do not indicate temperature increase or decrease. Note that each of the cells C00, C10, C20, ..., C24, C34, C44 is composed of 108 (=12 × 9) pixels.

[0024] A specific example of the first noise reduction process will be explained with reference to Figure 3. In this example, the lower limit was set to 3.

[0025] Cell C00, which constitutes the first difference image P1, had 15 pixels indicating temperature increase (grayed pixels) and 12 pixels indicating temperature decrease (hatched pixels). Both 15 and 12 are greater than the lower limit of 3. Therefore, the first noise reduction processing unit 22 does not perform the process of "changing all pixels in cell C00 to pixels that do not indicate temperature increase or decrease." In other words, the temperature changes shown by each pixel in cell C00 that constitutes the first difference image P1 are carried over directly to cell C00 that constitutes the second difference image P2 (see Figure (A)).

[0026] Cell C22, which constitutes the first difference image P1, had 108 pixels indicating temperature increase and 0 pixels indicating temperature decrease. 0 is less than the lower limit of 3. Therefore, the first noise reduction processing unit 22 performs a process to change all pixels constituting cell C22 to pixels that do not indicate either temperature increase or temperature decrease (see Figure (B)).

[0027] Cell C44, which constitutes the first difference image P1, had 0 pixels indicating temperature increase and 91 pixels indicating temperature decrease. 0 is less than the lower limit of 3. Therefore, the first noise reduction processing unit 22 performs a process to change all pixels constituting cell C44 to pixels that do not indicate either temperature increase or temperature decrease (see Figure (C)).

[0028] If the cells constituting the first difference image P1 contain both pixels indicating a rise in temperature and pixels indicating a fall in temperature, it is considered that this rise and fall in temperature occurred when pedestrians W1 and W2 passed through the portion of detection area A corresponding to those cells. On the other hand, if the cells constituting the first difference image P1 contain only one of either pixels indicating a rise in temperature or pixels indicating a fall in temperature, it is considered that this rise or fall in temperature occurred due to an external factor. The first noise reduction process allows us to obtain a second difference image P2 that records only the temperature changes that are thought to have occurred due to the passage of pedestrians W1 and W2.

[0029] The measurement unit 23 measures the number of pedestrians W1 that passed through detection area A in the M1 direction and the number of pedestrians W2 that passed through in the M2 direction, based on the second difference image P2 generated by the first noise reduction processing unit 22. Various known methods can be used for this measurement.

[0030] The external interface unit 14 is connected to a suitable external device. The external interface unit 14 can transmit the measurement results from the measurement unit 23 to the external device at predetermined timings and in predetermined formats.

[0031] The audio output unit 15 consists of a buzzer or a speaker. The audio output unit 15 can inform people in the vicinity of the measurement results by outputting audio corresponding to the measurement results from the measurement unit 23.

[0032] The light-emitting unit 16 is composed of at least one light-emitting diode. The light-emitting unit 16 can notify people in the vicinity of the measurement results by lighting up or flashing in a manner (color, brightness, etc.) corresponding to the measurement results from the measurement unit 23.

[0033] The setting unit 13 consists of DIP switches. The user can set the aforementioned lower limit and the maximum ratio described later via the setting unit 13. The user can also individually switch the external interface unit 14, the audio output unit 15, and the light emission unit 16 on or off via the setting unit 13.

[0034] Figure 4 is an operation flowchart of the passerby counting device 10A according to this embodiment, which describes the movements of each part described above in chronological order.

[0035] In step S1-1, the storage unit 12 stores the first thermal image generated by the thermal image generation unit 11.

[0036] In step S1-2, which is executed 1 / 10 s after step S1-1, the storage unit 12 stores the second thermal image generated by the thermal image generation unit 11.

[0037] In step S1-3, which is performed after step S1-2, the difference image generation unit 21 generates a first difference image P1 from the first thermal image and the second thermal image.

[0038] In step S1-4, which is performed after step S1-3, the first noise reduction processing unit 22 generates a second difference image P2 by applying a first noise reduction process to the first difference image P1.

[0039] In step S1-5, which is performed after step S1-4, the measurement unit 23 measures the number of people passing through based on the second difference image P2.

[0040] The person-counting device 10A in this embodiment operates repeatedly according to this operation flow every 1 / 10 second. In the second cycle of the flow, the second thermal image generated and stored in the first cycle becomes the first thermal image. The same applies to the third cycle and beyond.

[0041] Thus, in the person-passing-person measuring device 10A according to this embodiment, the number of people passing by is measured based on a second difference image P2 from which the influence of external factors that cause noise has been removed. For this reason, the person-passing-person measuring device 10A can accurately measure the number of people passing by without being affected by external factors.

[0042] [Second Example] Figure 5 shows a passerby counting device 10B according to a second embodiment of the present invention. The passerby counting device 10B differs from the passerby counting device 10A in that it is equipped with a control unit 20B instead of the control unit 20A, but is otherwise common to the passerby counting device 10A.

[0043] The control unit 20B, like the control unit 20A, is composed of a microprocessor and a computer program executed on that processor. However, in this embodiment, in addition to the difference image generation unit 21, the first noise reduction processing unit 22, and the measurement unit 23, a determination processing unit 24 and a second noise reduction processing unit 25 are formed within the control unit 20B by the execution of the computer program.

[0044] The determination processing unit 24 performs a determination process to determine whether or not the first noise reduction processing by the first noise reduction processing unit 22 is necessary, based on the first difference image P1 generated by the difference image generation unit 21. Here, the determination process involves measuring the number of pixels indicating temperature rise and the number of pixels indicating temperature fall in the entire first difference image P1, and determining that the first noise reduction processing is unnecessary if either of the two measured numbers is smaller than a predetermined lower limit, and otherwise determining that the first noise reduction processing is necessary. The determination process can also be described as a process to confirm that pedestrians W1 and W2 are clearly not present within the detection area A.

[0045] The second noise reduction processing unit 25 generates a third difference image P3 (more precisely, data relating to the third difference image P3) by applying the second noise reduction processing to the first difference image P1 generated by the difference image generation unit 21, only when the determination processing unit 24 determines that it is unnecessary (i.e., when there are clearly no pedestrians W1 and W2 passing through detection area A). Here, the second noise reduction processing refers to the process of changing all pixels constituting the first difference image P1 to pixels that do not indicate either temperature increase or decrease.

[0046] If the determination processing unit 24 determines that it is necessary (i.e., if there may be pedestrians W1 and W2 passing through detection area A), the first noise reduction processing unit 22 generates a second difference image P2 by applying a first noise reduction process to the first difference image P1, similar to the first embodiment.

[0047] Referring to Figure 6, specific examples of the first and second noise reduction processes will be explained. In this example, the lower limit was set to 1.

[0048] If the first difference image P1, which consists of 2700 pixels, contains both pixels indicating a number of temperature increases above a lower limit and pixels indicating a number of temperature decreases above a lower limit, the first noise reduction processing unit 22 performs a first noise reduction process on the first difference image P1, and as a result, the second difference image P2 is generated (see Figure (A)).

[0049] On the other hand, if the first difference image P1 does not contain any pixels indicating a decrease in temperature of a number greater than or equal to the lower limit, the second noise reduction processing unit 25 applies a second noise reduction process to the first difference image P1, and as a result, the third difference image P3 is generated (see Figure (B)). Similarly, if the first difference image P1 does not contain any pixels indicating an increase in temperature of a number greater than or equal to the lower limit, the second noise reduction processing unit 25 applies a second noise reduction process to the first difference image P1, and as a result, the third difference image P3 is generated (see Figure (C)).

[0050] The measurement unit 23 measures the number of pedestrians W1 that passed through detection area A in the M1 direction and the number of pedestrians W2 that passed through in the M2 direction, based on the second difference image P2 generated by the first noise reduction processing unit 22 or the third difference image P3 generated by the second noise reduction processing unit 25.

[0051] Figure 7 is an operation flowchart of the passerby counting device 10B according to this embodiment.

[0052] In step S2-1, the storage unit 12 stores the first thermal image generated by the thermal image generation unit 11.

[0053] In step S2-2, which is executed 1 / 10 s after step S2-1, the storage unit 12 stores the second thermal image generated by the thermal image generation unit 11.

[0054] In step S2-3, which is performed after step S2-2, the difference image generation unit 21 generates a first difference image P1 from the first thermal image and the second thermal image.

[0055] In step S2-4, which is performed after step S2-3, the determination processing unit 24 performs determination processing based on the first difference image P1.

[0056] In step S2-5, which is executed if it is determined in step S2-4 that a first noise reduction process is necessary, the first noise reduction processing unit 22 generates a second difference image P2 by applying the first noise reduction process to the first difference image P1.

[0057] In step S2-6, which is executed when it is determined in step S2-4 that the first noise reduction process is unnecessary, the second noise reduction processing unit 25 generates a third difference image P3 by applying the second noise reduction process to the first difference image P1.

[0058] In step S2-7, which is performed after step S2-5 or step S2-6, the measurement unit 23 measures the number of people passing through based on the second difference image P2 or the third difference image P3.

[0059] The person-counting device 10B according to this embodiment operates repeatedly according to this operation flow every 1 / 10 second. In the second cycle of the flow, the second thermal image generated and stored in the first cycle becomes the first thermal image. The same applies to the third cycle and beyond.

[0060] Thus, in the pedestrian counting device 10B according to this embodiment, the first noise reduction process, which has a greater processing load than the second noise reduction process, is executed only when there is a possibility that pedestrians W1 and W2 are passing through the detection area A. For this reason, the pedestrian counting device 10B can reduce the processing load on the control unit 20B and accurately measure the number of people passing through without being affected by external factors.

[0061] [Differentiation] Although the first and second embodiments of the person-counting device according to the present invention have been described above, the configuration of the present invention is not limited to these.

[0062] For example, the people passing through detection area A, which is set up near the entrances and exits of various facilities other than commercial facilities or various vehicles, may be used to measure people passing through the area.

[0063] Furthermore, the mounting location for the people-passing measurement devices 10A and 10B may be on a wall.

[0064] Furthermore, the first noise reduction process may involve measuring the number of pixels indicating temperature increase (for example, 12 pixels; the same applies hereafter) and the number of pixels indicating temperature decrease (2 pixels) in each of the multiple cells constituting the first difference image P1. If there is a cell in which the ratio of the larger of the two measured numbers to the smaller one (6 times) is greater than a predetermined maximum ratio, then all pixels constituting that cell may be changed to pixels that do not indicate temperature increase or decrease. An example of the maximum ratio is 10 times.

[0065] In this invention, the content of the first noise reduction process can be appropriately modified, as long as it is "a process in which the number of pixels indicating temperature increase and the number of pixels indicating temperature decrease in each of the multiple cells constituting the first difference image P1 are measured, and if there is a cell in which the two measured numbers or their ratio satisfy a predetermined determination condition, all pixels constituting that cell are changed to pixels that do not indicate temperature increase or decrease."

[0066] Alternatively, the determination process may involve measuring the number of pixels indicating temperature increase (e.g., 1000 pixels; the same applies hereafter) and the number of pixels indicating temperature decrease (4 pixels) in the entire first difference image P1, and determining that the first noise reduction process is unnecessary if the ratio of the larger of the two measured numbers to the smaller one (250 times) is greater than a predetermined maximum ratio. An example of the maximum ratio is 200 times.

[0067] In this invention, the content of the determination process can be appropriately modified, as long as it is "a process that measures the number of pixels indicating temperature rise and the number of pixels indicating temperature fall in the entirety of the first difference image P1, and determines that a first noise reduction process is necessary if the two measured numbers or their ratio satisfy predetermined determination conditions, and determines that a first noise reduction process is unnecessary otherwise."

[0068] Furthermore, if the people-passing measurement devices 10A and 10B can generate a thermal image (more precisely, data related to the thermal image) relating to the thermal change of detection area A that occurred during a predetermined time (for example, 1 / 10 s) using their thermal image generation unit 11, they do not need to be equipped with a difference image generation unit 21 (and storage unit 12). In this case, the first noise reduction processing unit 22 and the determination processing unit 24 use the thermal image generated by the thermal image generation unit 11 as the first difference image P1.

[0069] Furthermore, the people passing through measuring devices 10A and 10B do not necessarily have to include all or part of the setting unit 13, the external interface unit 14, the audio output unit 15, and the light-emitting unit 16.

[0070] The number of pixels (2700) and cells (25) constituting the first difference image P1, etc., in the first and second embodiments are merely examples. [Explanation of Symbols]

[0071] 10A, 10B Passing Person Counting Device 11 Thermal Image Generation Unit 12 Storage section 13. Settings Section 14 External Interface Section 15 Audio output section 16 Light-emitting part 20A, 20B Control Unit 21 Difference Image Generation Unit 22 First Noise Reduction Processing Unit 23 Measurement Unit 24. Determination Processing Unit 25. Second Noise Reduction Processing Unit 100 ceiling 200 beds P1 First difference image P2 Second difference image P3 Third Difference Image W1, W2 Pedestrians

Claims

1. A passerby counting device that measures the number of people passing through a predetermined detection area, A thermal image generation unit generates a thermal image of the detection area each time time Δt elapses, A difference image generation unit generates a first difference image, which is the difference image between the thermal image generated at time t and the thermal image generated at time t + Δt. A first noise reduction processing unit that applies a first noise reduction process to the first difference image, A measurement unit that measures the number of people who have passed through the detection area based on a second difference image, which is a difference image after the first noise reduction process, Equipped with, The first noise reduction process is a process that (1) measures the number of pixels indicating temperature increase and the number of pixels indicating temperature decrease in each of the multiple cells constituting the first difference image, and if there is a cell in which either of the two measured numbers is smaller than a predetermined lower limit, changes all the pixels constituting that cell to pixels that do not indicate temperature increase or temperature decrease, or (2) measures the number of pixels indicating temperature increase and the number of pixels indicating temperature decrease in each of the multiple cells constituting the first difference image, and if there is a cell in which the ratio of the larger of the two measured numbers to the smaller one is greater than a predetermined maximum ratio, changes all the pixels constituting that cell to pixels that do not indicate temperature increase or temperature decrease. A device for measuring the number of people passing through, characterized by the following features.

2. A passerby counting device that measures the number of people passing through a predetermined detection area, A thermal image generation unit generates a first difference image relating to the thermal change of the detection area that occurred during a predetermined period of time, A first noise reduction processing unit that applies a first noise reduction process to the first difference image, A measurement unit that measures the number of people who have passed through the detection area based on a second difference image, which is a difference image after the first noise reduction process, Equipped with, The first noise reduction process is a process that (1) measures the number of pixels indicating temperature increase and the number of pixels indicating temperature decrease in each of the multiple cells constituting the first difference image, and if there is a cell in which either of the two measured numbers is smaller than a predetermined lower limit, changes all the pixels constituting that cell to pixels that do not indicate temperature increase or temperature decrease, or (2) measures the number of pixels indicating temperature increase and the number of pixels indicating temperature decrease in each of the multiple cells constituting the first difference image, and if there is a cell in which the ratio of the larger of the two measured numbers to the smaller one is greater than a predetermined maximum ratio, changes all the pixels constituting that cell to pixels that do not indicate temperature increase or temperature decrease. A device for measuring the number of people passing through, characterized by the following features.

3. A passerby counting device that measures the number of people passing through a predetermined detection area, A thermal image generation unit generates a thermal image of the detection area each time time Δt elapses, A difference image generation unit generates a first difference image, which is the difference image between the thermal image generated at time t and the thermal image generated at time t + Δt. A determination processing unit performs a determination process to determine whether or not a first noise reduction process is necessary based on the first difference image, A first noise reduction processing unit that performs the first noise reduction processing on the first difference image when it is determined that the first noise reduction processing is necessary, If it is determined that the first noise reduction process is unnecessary, a second noise reduction processing unit performs a second noise reduction process on the first difference image, A measurement unit that measures the number of people who have passed through the detection area based on a second difference image, which is the difference image after the first noise reduction process, or a third difference image, which is the difference image after the second noise reduction process. Equipped with, The determination process is a process that (1) measures the number of pixels indicating temperature increase and the number of pixels indicating temperature decrease in the entire first difference image, and determines that the first noise reduction process is unnecessary if either of the two measured numbers is smaller than a predetermined lower limit, and determines that the first noise reduction process is necessary otherwise, or (2) measures the number of pixels indicating temperature increase and the number of pixels indicating temperature decrease in the entire first difference image, and determines that the first noise reduction process is unnecessary if the ratio of the larger of the two measured numbers to the smaller one is greater than a predetermined maximum ratio, and determines that the first noise reduction process is necessary otherwise. The first noise reduction process is a process that (1) measures the number of pixels indicating temperature increase and the number of pixels indicating temperature decrease in each of the multiple cells constituting the first difference image, and if there is a cell in which either of the two measured numbers is smaller than a predetermined lower limit, changes all the pixels constituting that cell to pixels that do not indicate temperature increase or temperature decrease, or (2) measures the number of pixels indicating temperature increase and the number of pixels indicating temperature decrease in each of the multiple cells constituting the first difference image, and if there is a cell in which the ratio of the larger of the two measured numbers to the smaller one is greater than a predetermined maximum ratio, changes all the pixels constituting that cell to pixels that do not indicate temperature increase or temperature decrease. The second noise reduction process is a process that changes all pixels constituting the first difference image to pixels that do not show either temperature increase or temperature decrease. A device for measuring the number of people passing through, characterized by the following features.

4. A passerby counting device that measures the number of people passing through a predetermined detection area, A thermal image generation unit generates a first difference image relating to the thermal change of the detection area that occurred during a predetermined period of time, A determination processing unit performs a determination process to determine whether or not a first noise reduction process is necessary based on the first difference image, A first noise reduction processing unit that performs the first noise reduction processing on the first difference image when it is determined that the first noise reduction processing is necessary, If it is determined that the first noise reduction process is unnecessary, a second noise reduction processing unit performs a second noise reduction process on the first difference image, A measurement unit that measures the number of people who have passed through the detection area based on a second difference image, which is the difference image after the first noise reduction process, or a third difference image, which is the difference image after the second noise reduction process. Equipped with, The determination process is a process that (1) measures the number of pixels indicating temperature increase and the number of pixels indicating temperature decrease in the entire first difference image, and determines that the first noise reduction process is unnecessary if either of the two measured numbers is smaller than a predetermined lower limit, and determines that the first noise reduction process is necessary otherwise, or (2) measures the number of pixels indicating temperature increase and the number of pixels indicating temperature decrease in the entire first difference image, and determines that the first noise reduction process is unnecessary if the ratio of the larger of the two measured numbers to the smaller one is greater than a predetermined maximum ratio, and determines that the first noise reduction process is necessary otherwise. The first noise reduction process is a process that (1) measures the number of pixels indicating temperature increase and the number of pixels indicating temperature decrease in each of the multiple cells constituting the first difference image, and if there is a cell in which either of the two measured numbers is smaller than a predetermined lower limit, changes all the pixels constituting that cell to pixels that do not indicate temperature increase or temperature decrease, or (2) measures the number of pixels indicating temperature increase and the number of pixels indicating temperature decrease in each of the multiple cells constituting the first difference image, and if there is a cell in which the ratio of the larger of the two measured numbers to the smaller one is greater than a predetermined maximum ratio, changes all the pixels constituting that cell to pixels that do not indicate temperature increase or temperature decrease. The second noise reduction process is a process that changes all pixels constituting the first difference image to pixels that do not show either temperature increase or temperature decrease. A device for measuring the number of people passing through, characterized by the following features.

Images