JP7583273B2 - Hand-washing recognition device, hand-washing recognition method, hand-washing recognition program, and hand-washing recognition system - Google Patents

Description

The present invention relates to a hand-washing recognition device, a hand-washing recognition method, a hand-washing recognition program, and a hand-washing recognition system.

In recent years, the importance of hand washing as an infection prevention measure has been reassessed due to the spread of viruses and pathogens. For example, hygiene management for workers is particularly important in food handling sites, hospitals, and nursing homes, and hand washing is common practice. In particular, there is the international standard HACCP for hygiene management records that targets food-related businesses, and it is required to check, monitor, and record hygiene management actions.

Hand-washing technologies are disclosed below.

JP 2016-211268 A JP 2020-048628 A JP 2020-091739 A

However, when washing your hands, for example, the surfactant action of soap lifts bacteria and dirt from the surface of your skin, trapping them in the soap bubbles. In other words, the soap bubbles left after washing your hands contain a large amount of bacteria and dirt. Therefore, if the soap bubbles are not washed off thoroughly (by rinsing the bubbles off your hands with running water), bacteria and dirt will remain on the surface of your hands. Also, if the soap bubbles are not washed off thoroughly, rubbing your hands together can spread bacteria and dirt, making your hands less hygienic than before you washed them.

The present disclosure provides a hand-washing recognition device, a hand-washing recognition method, a hand-washing recognition program, and a hand-washing recognition system that determine whether washing has been performed properly.

The device has an image unit that acquires an image of the hand washing area, a first detection unit that detects a person's hands from the image, a monitoring unit that monitors an estimated foam amount that indicates the amount of foam estimated to be adhering to the hands, a second detection unit that detects the process of wetting the hands of the person when washing the hands, and a determination unit that sets a first threshold based on the estimated foam amount in the process of wetting the hands and determines that the foam has been properly rinsed off when the estimated foam amount remains below the first threshold for a first period of time.

The present disclosure provides a hand washing recognition device, a hand washing recognition method, a hand washing recognition program, and a hand washing recognition system that determine whether washing has been performed properly during hand washing.

FIG. 1 is a diagram showing an example of the configuration of a hand-washing recognition system 10. FIG. 2 is a diagram illustrating an example of the configuration of the image processing device 100. FIG. 3 is a diagram showing an example of a process flowchart of the hand-washing determination process S100. FIG. 4 is a diagram showing an example of calculation of an estimated foam volume. FIG. 5 is a diagram showing an example of calculation of an estimated amount of lather when hands are wet. FIG. 6 is a diagram showing an example of a change in the estimated foam volume during hand washing. FIG. 7 is a diagram showing an example of hand movements during hand washing. FIG. 8 is a diagram showing an example in which the determination region T10 is a polygon. FIG. 9 is a diagram showing an example of a hand-washing process.

[First embodiment]
A first embodiment will be described.

<Configuration example of hand washing recognition system 10>
1 is a diagram showing an example of the configuration of a hand-washing recognition system 10. The hand-washing recognition system 10 includes an image processing device (hand-washing recognition device) 100 and a camera 200. The hand-washing recognition system 10 analyzes image data of a video captured by the camera 200, and judges whether or not a target person has properly rinsed their hands.

Camera 200 is a photographing device that photographs the hands of a target person at a washbasin. Camera 200 is installed in a place where a person washes their hands, such as a toilet or kitchen, and photographs the hands of the target person with a focus on the washbasin, such as a sink. Camera 200 is connected to image processing device 100, for example, via a wired connection (such as a network cable) or wireless connection (such as Bluetooth (registered trademark)), and transmits the photographed image data to image processing device 100. Camera 200, for example, photographs a video as an image, and continuously transmits video data of a predetermined data size or a predetermined time to image processing device 100. Camera 200 also photographs video constantly or periodically. Camera 200 may also photograph a series of still images instead of a video.

Target person 300 is a person who is washing his/her hands and is the person for whom it is determined whether or not the washing has been performed properly. Target person 300 washes his/her hands with soap at a hand washing area.

The image processing device 100 is a device that analyzes the video data acquired from the camera 200 and determines whether the target person 300 shown in the video data has been properly washed off, and is, for example, a computer or a server machine.

In the hand-washing recognition system 10, the camera 200 transmits captured image data to the image processing device 100. The image processing device 100 detects that the hand of the target person 300 is captured in the image data. The image processing device 100 measures (calculates) the amount of foam attached to the hand (estimated foam amount), and when the amount of foam satisfies a rinsing completion condition, determines that the target person 300 has been properly rinsed. The rinsing completion condition is, for example, that the estimated foam amount (including the average and maximum estimated foam amount) is equal to or less than a threshold value for a predetermined time (first time). The image processing device 100, for example, regards the white portion in the hand area as foam (estimated foam) and calculates the estimated foam amount.

The image processing device 100 detects the state in which the target person 300 has wet hands before applying soap, and calculates an estimated amount of lather (initial value) for wet hands. In reality, even when the hands are merely wet with water before applying soap, parts of the hands may appear white due to light reflection, and this white part may be mistaken for lather. Therefore, the image processing device 100 measures the estimated amount of lather on wet hands as an initial value, and uses this to set a threshold value, thereby improving the accuracy of determining the amount of lather remaining on the hands.

<Configuration example of image processing device 100>
2 is a diagram illustrating an example of the configuration of the image processing device 100. The image processing device 100 includes a CPU (Central Processing Unit) 110, a storage 120, a memory 130, a communication circuit 140, and an accelerator 150.

Storage 120 is an auxiliary storage device such as a flash memory, a hard disk drive (HDD), or a solid state drive (SSD) that stores programs and data. Storage 120 stores a hand-washing determination program 121.

Memory 130 is an area into which programs stored in storage 120 are loaded. Memory 130 may also be used as an area into which programs store data.

The CPU 110 is a processor that loads programs stored in the storage 120 into the memory 130, executes the loaded programs, constructs each part, and realizes each process.

The communication circuit 140 is a circuit that communicates with the camera 200. For example, the communication circuit 140 is a network interface card (NIC). The communication circuit 140 is also a Bluetooth connection adapter. The communication means in the communication circuit 140 may be either wireless or wired.

The accelerator 150 is, for example, a GPU (Graphics Processing Unit) or an FPGA (Field Programmable Gate Array) that performs specific processing according to instructions from the CPU 110. The accelerator 150 is, for example, a hardware accelerator that performs image processing, arithmetic processing, etc.

By executing the hand-washing judgment program 121, the CPU 110 constructs an image section, a first detection section, a second detection section, a monitoring section, and a judgment section, and performs hand-washing judgment processing. The hand-washing judgment processing is processing for judging whether the target person 300 has washed his/her hands properly.

In the hand-washing determination process, the image processing device 100 acquires image data from the camera 200. Then, the image processing device 100 detects the hands of the target person 300 from the image data. Then, the image processing device 100 detects from the image data that the target person 300 is performing the process of wetting his/her hands. Then, the image processing device 100 calculates an initial value of the estimated amount of foam in the process in which the target person 300 wets his/her hands, and sets a rinse-off threshold (first threshold) based on the initial value. Furthermore, the image processing device 100 monitors the estimated amount of foam attached to the hands, and when the rinse-off completion condition is met, determines that the target person 300 has properly rinsed his/her hands.

The CPU 110 executes the hand detection module 1211 of the hand-washing determination program 121 to construct a first detection unit and perform hand detection processing. The hand detection processing is processing for detecting a person's hand (right hand, left hand, etc.) from image data. The image processing device 100 detects hands from an image using, for example, a convolutional neural network.

The CPU 110 executes the hand wetting process detection module 1212 of the hand washing determination program 121 to construct a second detection unit and perform hand wetting process detection processing. The hand wetting process detection processing is processing for detecting, from image data, whether a target person is (has) performed the process of wetting their hands.

The CPU 110 executes the initial lather volume calculation module 1213 of the hand-washing determination program 121 to construct a determination unit and perform an initial lather volume calculation process. The initial lather volume calculation process is a process for calculating an initial estimated lather volume for the hands when it is detected from the image data that the subject person is (has) performed the process of wetting the hands. Note that in the initial lather volume calculation process, the image processing device 100 calculates, for example, an initial estimated lather volume for each of the right hand and the left hand.

The CPU 110 executes the lather volume monitoring module 1214 of the hand-washing determination program 121 to construct a monitoring section and perform lather volume monitoring processing. The lather volume monitoring processing is a process of monitoring the estimated lather volume on the hands after detecting from image data that the subject person is (has) undergone the process of wetting the hands. In the lather volume monitoring processing, the image processing device 100 monitors the estimated lather volume of each of the right hand and the left hand.

The CPU 110 executes the rinsing determination module 1215 of the hand-washing determination program 121 to construct a determination unit and perform rinsing determination processing. The rinsing determination processing is a process in which the estimated amount of foam is compared with a rinsing completion condition (a condition for determining that rinsing has been performed appropriately) to determine whether or not the rinsing completion condition is met. If the rinsing completion condition is met, the image processing device 100 determines that the target person 300 has been properly rinsed. The rinsing completion condition is, for example, that the estimated amount of foam is equal to or less than a rinsing threshold for a predetermined period of time.

<About hand washing judgment processing>
The hand-washing determination process S100 will now be described with reference to FIG 3, which is a diagram showing an example of a processing flowchart of the hand-washing determination process S100.

The image processing device 100 starts acquiring images in the hand-washing determination process S100 (S100-1). The images are, for example, image data captured by the camera 200, and are acquired periodically from the camera 200. The image data are, for example, video data.

The image processing device 100 waits to detect the target person's hand from the image (No in S100-2). When the image processing device 100 detects the target person's hand from the image (Yes in S100-2), it waits to detect the process of wetting the hand (No in S100-3). The method of detecting the process of wetting the hand will be described later. When the image processing device 100 detects the process of wetting the hand (Yes in S100-3), it calculates an initial value for the estimated amount of lather (S100-4).

Figure 4 is a diagram showing an example of calculating an estimated amount of foam. Figure 4 (A) is a diagram showing an example of an image at the time when foam is taken on the hand. Figure 4 (B) is a diagram showing an example of a hand region. Figure 4 (C) is a diagram showing an example of a foam region. The image processing device 100 detects the hand from the image of Figure 4 (A) and detects the hand region T1 of Figure 4 (B). Then, the image processing device 100 detects the white component in the hand region T1 as foam from the image of Figure 4 (A) and detects the foam region T2 of Figure 4 (C). Then, the image processing device 100 calculates the ratio of the area of the foam region T2 to the area of the hand region T1 as the estimated amount of foam.

Figure 5 is a diagram showing an example of calculation of the estimated amount of foam when the hands are wet. As shown in Figure 5, when the process of wetting the hands is performed, parts of the hands may appear white due to factors other than foam, such as light reflection. Since the image processing device 100 determines the white components in the hand area as estimated foam, it may also determine areas T3 and T4 that appear white due to light reflection as estimated foam. Therefore, the image processing device 100 calculates an initial value of the estimated amount of foam from the time the hands are wetted until the foam is taken up by the hands. The initial value is a numerical value indicating the amount (proportion) of estimated foam at the time the hands are wetted. The image processing device 100 uses the initial value of the estimated amount of foam to determine whether rinsing has been performed appropriately. The initial value includes, for example, the maximum or average value of the estimated amount of foam within a specified time.

Returning to the processing flowchart of FIG. 3, the image processing device 100 sets the wash-off threshold based on the calculated initial value (S100-5). The wash-off threshold is a threshold used as a wash-off completion condition, and is a threshold for determining whether or not the foam has been sufficiently washed away. The image processing device 100 sets the initial value, or a value obtained by increasing or decreasing the initial value by a few percent, as the wash-off threshold.

The image processing device 100 starts monitoring the estimated amount of foam on the hand (S100-6). During monitoring, the image processing device 100 calculates the estimated amount of foam in the hand area at each predetermined timing.

At the timing when the estimated foam volume is calculated, the image processing device 100 determines whether the estimated foam volume satisfies a rinse-off completion condition (S100-7). The rinse-off completion condition may be, for example, that the estimated foam volume is equal to or less than the rinse-off threshold for a predetermined time. Alternatively, the rinse-off completion condition may be, for example, that the average estimated foam volume for a predetermined time is equal to or less than the rinse-off threshold.

Figure 6 shows an example of the change in the estimated amount of foam during hand washing. The estimated amount of foam increases during the process of wetting the hands. The image processing device 100 sets the value near the maximum increase as the initial value.

The estimated amount of foam gradually increases during the process of lathering the soap. Note that the estimated amount of foam is an estimate of the amount of foam attached to the hand, so it also increases or decreases depending on the direction or angle of the hand. Because hands are frequently moved during the process of lathering soap, the estimated amount of foam may increase or decrease repeatedly, but generally shows a tendency to increase over time.

The estimated foam volume decreases during the rinsing process. The estimated foam volume decreases to close to the initial value. During the rinsing process, when most of the foam has been rinsed away, the hands are left with water. The estimated foam volume in this state is close to the initial value measured during the wetting the hands process.

As shown in FIG. 6, the image processing device 100 determines that rinsing has been performed appropriately when the estimated foam amount becomes less than the rinse threshold set from the initial value based on the approximation between the estimated foam amount (initial value) during the process of wetting the hands and the estimated foam amount after rinsing is complete. Note that the image processing device 100 takes into account measurement errors in the estimated foam amount and sets the condition for rinsing completion to be that the estimated foam amount is less than the rinse threshold for a predetermined consecutive period of time. The image processing device 100 may also take into account measurement errors in the estimated foam amount and set the condition for rinsing completion to be that the average estimated foam amount for a predetermined period of time is less than the rinse threshold.

Returning to the processing flowchart of FIG. 3, if the estimated foam volume satisfies the rinsing completion condition (Yes in S100-7), the image processing device 100 determines that rinsing has been performed appropriately (S100-8) and ends the processing.

On the other hand, if the estimated foam volume does not satisfy the rinsing completion condition (No in S100-7), the image processing device 100 judges whether or not a hand is captured in the image (S100-9). By detecting the hand, the image processing device 100 judges whether or not the target person is continuing to wash their hands.

If the image processing device 100 cannot detect hands (No in S100-9), it assumes that the subject person has finished washing their hands, determines that rinsing was inappropriate (S100-10), and ends the process.

On the other hand, if the image processing device 100 detects hands (Yes in S100-9), it assumes that the target person is still washing their hands and judges whether the rinsing completion condition is met (S100-7). In this way, as long as the target person continues to wash their hands, the image processing device 100 continues to monitor the estimated amount of foam until the rinsing completion condition is met or until hands can no longer be detected.

<Conditions for completing washing>
The estimated amount of lather is calculated, for example, by integrating both hands. The image processing device 100 calculates the estimated amount of lather as the total area of the right hand region and the left hand region, or the amount of lather relative to the area of the overlapping hand regions of both hands. The image processing device 100 may determine that the rinsing has been properly performed based on the estimated amount of lather for both hands as the condition for completing the rinsing. The image processing device 100 may determine that the rinsing has been properly performed in a series of hand washing operations, for example, when the total value of the estimated amount of lather for the right hand and the left hand falls below a rinsing threshold.

The estimated amount of lather may also be calculated, for example, for each of the right and left hands. The image processing device 100 calculates the amount of lather relative to the area of the hand region of the right hand as the estimated amount of lather for the right hand. The image processing device 100 calculates the amount of lather relative to the area of the hand region of the left hand as the estimated amount of lather for the left hand. The image processing device 100 may calculate the estimated amount of lather for each of both hands, and the condition for completion of rinsing may be that both hands have been properly rinsed. The image processing device 100 may determine, for example, whether or not rinsing has been properly performed for each of the right and left hands, and may determine that both hands have been properly rinsed as a condition for determining that rinsing in a series of hand washing has been properly performed.

<Detection of the process of wetting hands>
A method for detecting the process of wetting the hands will be described. The method for detecting the process of wetting the hands includes, for example, a method for detecting based on the movement of the hands (first method) and a method for detecting by detecting running water (second method). Each method will be described below.

<1. Method of detection based on hand movements (first method)>
7 is a diagram showing an example of hand movements during hand washing. A flow line L1 is a line showing an example of hand movements.

When the target person starts washing their hands, their hands appear in area A1 outside the sink. The hands move along line of movement L1 and stay in area A2 within judgment area T10 for a short period of time, from one second to a few seconds. At this time, the target person is in the process of wetting their hands. Judgment area T10 is an area in which the process of wetting the hands and the process of rinsing the hands are assumed to be taking place, and is, for example, an area set with the area under running water as its center. Then, when the process of wetting the hands is completed, the hands move along line of movement L1 into area A3 outside judgment area T10, for example to pick up soap.

When the hand is located within the determination area T10, the image processing device 100 determines that the hand wetting process is being performed. This allows the image processing device 100 to reduce erroneous determinations that the hand wetting process is being performed, for example, even when the hand is in a position where running water does not hit the hand. Note that the hand also moves into the determination area T10 during the rinsing process. Thus, the image processing device 100 determines that the hand wetting process has been performed (is being performed) when it detects that the hand remains within the determination area T10 for a predetermined time (second time) within a few seconds from the start of hand washing (for example, when the hand is detected for the first time) or when it is detected for the first time since the start of hand washing. The image processing device 100 detects the hand from the image and monitors the hand movement, for example, using a convolutional neural network.

Whether or not the hand is located within the judgment region T10 is determined, for example, by determining whether or not the center of gravity (center) position of the hand is located within the judgment region T10. Whether or not the hand is located within the judgment region T10 may also be determined, for example, by determining whether or not a part of the hand region from the fingertips to the wrist is located within the judgment region T10.

In addition, in FIG. 7, the determination area T10 is rectangular, but the determination area T10 is not limited to a rectangle. FIG. 8 is a diagram showing an example in which the determination area T10 is polygonal. The determination area T10 is set based on the range in which the hand may be located when the hand is moved under running water from the fingertips to the wrist.

2. Method of detection by detecting flowing water (Second method)
The image processing device 100 detects that the process of wetting the hands has been performed by detecting running water (the start of running water). For example, in the case of a running water system that automatically runs water using a sensor, the detection of running water is realized by receiving a notification from the running water system that running water has started in conjunction with the sensor.

Furthermore, the detection of running water may be realized, for example, by recognizing running water from an image. The image processing device 100 determines that the process of wetting the hands has been performed a predetermined time (third time) after the detection of running water. The predetermined time is assumed to be the time during which the process of wetting the hands is performed, and is a short time such as one second. The image processing device 100 detects running water from an image, for example, using a convolutional neural network.

The process of wetting the hands may be detected using both the first method and the second method. For example, the image processing device 100 may consider detection by both the first method and the second method as detection of the process of wetting the hands. The image processing device 100 may also consider detection by at least one of the first method and the second method as detection of the process of wetting the hands.

[Other embodiments]
Other embodiments will be described. Fig. 9 is a diagram showing an example of a process of washing hands. For example, when a target person starts washing hands, as shown in Fig. 9, the target person first performs a process of wetting the hands. In the process of wetting the hands, running water is performed. Then, in the process of lathering soap, running water is stopped. Furthermore, in the process of rinsing, running water is performed again. The image processing device 100 needs to distinguish between the process of wetting the hands under running water and the rinsing process so as not to erroneously recognize it.

For example, the image processing device 100 excludes the process of wetting the hands from the monitoring by not monitoring the estimated foam volume for a predetermined time after detecting the initial value (rinsing threshold).

In addition, in a system where the water is flushed automatically (by a sensor, etc.), for example, the image processing device 100 may detect the running water during lathering. In this case, the image processing device 100 can detect the running water during the wetting of the hands process, detect that the running water has stopped during lathering, and then detect the running water during the rinsing process. The image processing device 100 may start monitoring the estimated foam volume when it detects the running water for rinsing (the second time the water is flushed in a series of hand washing).

To summarise the above, the following is added:

(Appendix 1)
An image unit for acquiring an image of a washroom;
a first detection unit that detects a person's hand from the image;
A monitoring unit for monitoring an estimated amount of foam, which indicates an amount of foam estimated to be attached to the hand;
a second detection unit that detects a process of wetting the person's hands while washing their hands;
A hand-washing recognition device having a judgment unit that sets a first threshold based on the estimated amount of foam during the hand wetting process, and judges that the foam has been properly rinsed off during the hand-washing process when the estimated amount of foam remains below the first threshold for a first period of time.

(Appendix 2)
The hand-washing recognition device according to claim 1, wherein the second detection unit determines that the hand wetting process is being performed when the time during which the hand is located within the determination area has exceeded a second time.

(Appendix 3)
The hand-washing recognition device according to claim 2, wherein the determination area includes a range that is set according to a position where water is running in the hand-washing area.

(Appendix 4)
The hand-washing recognition device according to claim 1, wherein the second detection unit determines that the hand-wetting process is being performed three hours after detecting running water in the hand-washing area.

(Appendix 5)
The hand-washing recognition device according to claim 4, wherein the second detection unit detects the running water by acquiring information that running water has started from an automatic water running system.

(Appendix 6)
The hand-washing recognition device according to claim 4, wherein the second detection unit detects the running water from the image.

(Appendix 7)
The hand-washing recognition device according to claim 1, wherein the monitoring unit determines a white component located within the hand area as the estimated foam.

(Appendix 8)
Acquire an image of the washroom,
Detecting a person's hands from the image;
monitoring an estimated amount of lather indicative of an estimated amount of lather adhering to said hands;
Detecting a process of wetting the person's hands during hand washing;
A hand washing recognition method, comprising: setting a first threshold value based on the estimated amount of foam during the hand wetting process; and determining that foam has been properly rinsed off during hand washing when the estimated amount of foam remains below the first threshold value for a first period of time.

(Appendix 9)
acquiring an image of a hand washing area;
detecting a person's hand from the image;
monitoring an estimated lather volume indicative of an estimated amount of lather estimated to be deposited on said hands;
detecting a process of wetting the person's hands during hand washing;
setting a first threshold based on the estimated amount of lather in the wetting step, and judging that lather has been properly rinsed off in the hand washing step when the estimated amount of lather is equal to or less than the first threshold for a first period of time;
A hand-washing recognition program that causes a computer to execute the above.

(Appendix 10)
A photographing device for photographing an image of a washroom;
acquiring the image from the imaging device;
Detecting a person's hands from the image;
monitoring an estimated amount of lather indicative of an estimated amount of lather adhering to said hands;
Detecting a process of wetting the person's hands during hand washing;
an image processing device that sets a first threshold based on the estimated amount of foam in the wetting of the hands, and determines that the foam has been properly rinsed off in the hand washing process when the estimated amount of foam is equal to or less than the first threshold for a first period of time;
A hand washing recognition system having

10: Hand-washing recognition system 100: Image processing device 110: CPU
120: Storage 121: Hand-washing determination program 1211: Hand detection module 1212: Hand-wetting process detection module 1213: Lather volume initial value calculation module 1214: Lather volume monitoring module 1215: Rinsing determination module 130: Memory 140: Communication circuit 150: Accelerator 200: Camera 300: Target person

Claims (8)

An image unit for acquiring an image of a washroom;
a first detection unit that detects a person's hand from the image;
A monitoring unit for monitoring an estimated amount of foam, which indicates an amount of foam estimated to be attached to the hand;
a second detection unit that detects a process of wetting the person's hands while washing their hands;
A hand-washing recognition device having a judgment unit that sets a first threshold based on the estimated amount of foam during the hand wetting process, and judges that the foam has been properly rinsed off during the hand-washing process when the estimated amount of foam remains below the first threshold for a first period of time. The hand-washing recognition device according to claim 1 , wherein the second detection unit determines that the hand wetting process is being performed when the time during which the hand is located within the determination area exceeds a second time. The hand-washing recognition device according to claim 2 , wherein the determination area includes a range that is set according to a position where water is flowing in the hand-washing area. The hand-washing recognition device according to claim 1 , wherein the second detection unit determines that the hand-wetting process is being performed three hours after detecting running water in the hand-washing area. The hand-washing recognition device according to claim 1 , wherein the monitoring unit determines a white component located within the hand area as the estimated foam. Acquire an image of the washroom,
Detecting a person's hands from the image;
monitoring an estimated amount of lather indicative of an estimated amount of lather adhering to said hands;
Detecting a process of wetting the person's hands during hand washing;
A hand washing recognition method, comprising: setting a first threshold value based on the estimated amount of foam during the hand wetting process; and determining that foam has been properly rinsed off during hand washing when the estimated amount of foam remains below the first threshold value for a first period of time. acquiring an image of a hand washing area;
detecting a person's hand from the image;
monitoring an estimated lather volume indicative of an estimated amount of lather estimated to be deposited on said hands;
detecting a process of wetting the person's hands during hand washing;
setting a first threshold based on the estimated amount of lather in the wetting step, and judging that lather has been properly rinsed off in the hand washing step when the estimated amount of lather is equal to or less than the first threshold for a first period of time;
A hand-washing recognition program that causes a computer to execute the above. A photographing device for photographing an image of a washroom;
acquiring the image from the imaging device;
Detecting a person's hands from the image;
monitoring an estimated amount of lather indicative of an estimated amount of lather adhering to said hands;
Detecting a process of wetting the person's hands during hand washing;
an image processing device that sets a first threshold based on the estimated amount of foam in the wetting of the hands, and determines that the foam has been properly rinsed off in the hand washing process when the estimated amount of foam is equal to or less than the first threshold for a first period of time;
A hand washing recognition system having

Images