Contamination Degree Estimation System, Contamination Degree Estimation Method, and Contamination Degree Estimation Program

Abstract

The present invention more accurately estimates a pollution level at a site subject to estimation.　A pollution level estimation system 10 for estimating a pollution level at a site subject to estimation comprises: an image acquisition unit 11 that acquires an image of the site subject to estimation; a detection unit 12 that detects a garbage portion and a non-garbage portion in the acquired image, the garbage portion showing garbage at the site subject to estimation, the non-garbage portion showing no garbage; and a pollution level estimation unit 13 that calculates the areas of the detected garbage portion and non-garbage portion, and estimates, on the basis of the calculated areas, the pollution level at the site subject to estimation.

Description

[Technical field]

[0001] The present invention relates to a pollution level estimation system, a pollution level estimation method, and a pollution level estimation program for estimating the pollution level of an estimation target location such as a coast. [Background technology]

[0002] The deterioration of coastal scenery has a negative impact on tourism and on the ecosystems of both marine and terrestrial organisms. For this reason, local governments clean up coastal garbage and request external organizations to carry out the cleanup work. At the same time, surveys are also conducted several times a year to grasp the current level of pollution of the coast. The surveys are conducted based on the guidelines of the Ministry of the Environment, which state that the coast is photographed with a camera and an investigator refers to the images to determine the level of pollution on an 11-point scale. Patent Document 1 also shows that investigators investigate drifting garbage on the coast. [Prior art documents] [Patent documents]

[0003] [Patent Document 1] JP 2003-143962 A Summary of the Invention [Problem to be solved by the invention]

[0004] The above-mentioned method of determining the degree of contamination based on images is subjective, and therefore does not necessarily provide an accurate evaluation. In addition, the method shown in Patent Document 1 is also based on visual inspection by an inspector, and therefore may not provide an accurate evaluation.

[0005] One embodiment of the present invention has been made in consideration of the above, and aims to provide a pollution level estimation system, a pollution level estimation method, and a pollution level estimation program that can more accurately estimate the pollution level of a location that is the subject of estimation. [Means for solving the problem]

[0006] In order to achieve the above-mentioned object, a contamination level estimation system according to one embodiment of the present invention is a contamination level estimation system that estimates the contamination level of a location that is to be estimated, and includes an image acquisition means that acquires an image showing the location that is to be estimated, a detection means that detects dust parts of the location that is to be estimated in the image acquired by the image acquisition means that show dust parts and non-dust parts that do not show dust, and a contamination level estimation means that calculates the area of ​​the dust parts and non-dust parts detected by the detection means, and estimates the contamination level of the location that is to be estimated based on the calculated areas.

[0007] In a contamination level estimation system according to an embodiment of the present invention, the contamination level of a target location is estimated based on the areas of dust and non-dust parts detected from an image. This makes it possible to estimate the contamination level without relying on the subjectivity of an investigator. Therefore, the contamination level estimation system according to an embodiment of the present invention can more accurately estimate the contamination level of a target location.

[0008] The contamination level estimation means acquires information indicating the distance from the imaging point to each of the dust and non-dust parts in the image, and calculates the areas of the dust and non-dust parts weighted by the distances. do. According to this configuration, the areas of the dust parts and non-dust parts used in estimating the contamination level can be determined taking the above-mentioned distance into consideration, thereby making it possible to more accurately estimate the contamination level of the location being estimated.

[0009] The contamination level estimation means may calculate the distance from the image acquired by the image acquisition means. With this configuration, it is possible to calculate the area weighted by the distance from only the image. As a result, it is possible to easily and accurately estimate the contamination level of the estimation target location.

[0010] The contamination level estimation means may calculate the number of pieces of dust captured in the location of the estimation target in the image and estimate the contamination level of the location of the estimation target based on the calculated number of pieces of dust. With this configuration, it is possible to estimate the contamination level of the location of the estimation target more accurately.

[0011] The detection means may detect dust parts, non-dust parts, and parts other than the location of the estimation target by segmentation using machine learning. With this configuration, it is possible to reliably and appropriately estimate the degree of contamination of the location of the estimation target.

[0012] Incidentally, one embodiment of the present invention can be described not only as an invention of a contamination level estimation system as described above, but also as an invention of a contamination level estimation method and a contamination level estimation program as described below. These are essentially the same invention, just in different categories, and have the same functions and effects.

[0013] That is, a contamination level estimation method according to one embodiment of the present invention is a contamination level estimation method which is an operating method of a contamination level estimation system which estimates the contamination level of a location that is an estimation target, and includes an image acquisition step of acquiring an image showing the location that is an estimation target, a detection step of detecting dust parts in which dust is shown and non-dust parts in which dust is not shown in the image acquired in the image acquisition step, and a contamination level estimation step of calculating an area of ​​the non-dust parts of the dust parts detected in the detection step and estimating the contamination level of the location that is an estimation target based on the calculated area. In the contamination level estimation step, information indicating the distance from the imaging point to each of the dust and non-dust parts in the image is obtained, and the areas of the dust and non-dust parts are calculated weighted by the distance. .

[0014] A contamination level estimation program according to one embodiment of the present invention is a contamination level estimation program that causes a computer to operate as a contamination level estimation system that estimates the contamination level of an estimation target location, and causes the computer to function as: an image acquisition means that acquires an image showing the estimation target location; a detection means that detects dust parts of the estimation target location in the image acquired by the image acquisition means that show dust and non-dust parts that do not show dust; and a contamination level estimation means that calculates the areas of the dust parts and non-dust parts detected by the detection means and estimates the contamination level of the estimation target location based on the calculated areas. The contamination level estimation means acquires information indicating the distance from the imaging point to each of the dust and non-dust parts in the image, and calculates the areas of the dust and non-dust parts weighted by the distances. . Effect of the Invention

[0015] According to one embodiment of the present invention, it is possible to more accurately estimate the pollution level of a location that is the subject of estimation. [Brief description of the drawings]

[0016] [Figure 1] 1 is a diagram showing a configuration of a contamination level estimation system according to an embodiment of the present invention. [Diagram 2] 1 is an example of an image used to estimate a contamination level and a detection result for the image. [Diagram 3] 3 is a flowchart showing a contamination level estimation method which is a process executed in the contamination level estimation system according to the embodiment of the present invention. [Figure 4] 1 is a diagram showing a configuration of a contamination level estimation program according to an embodiment of the present invention, together with a recording medium. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0017] Hereinafter, embodiments of a contamination level estimation system, a contamination level estimation method, and a contamination level estimation program according to the present invention will be described in detail with reference to the drawings. In the description of the drawings, the same elements are given the same reference numerals, and duplicated description will be omitted.

[0018] Fig. 1 shows a pollution level estimation system 10 according to this embodiment. The pollution level estimation system 10 is a system (apparatus) that estimates the pollution level of a location that is the subject of estimation. The pollution level in this embodiment is a degree that indicates the extent to which the location that is the subject of estimation is polluted by garbage. The pollution level may be estimated at multiple levels (e.g., five levels), for example.

[0019] In this embodiment, the location (area) of the estimation target is, for example, a coast. However, the location of the estimation target may be a location other than a coast as long as the above-mentioned pollution level can be estimated. For example, the location of the estimation target may be a river or an urban block. In this embodiment, an example will be described in which the location of the estimation target is a coast. When the location of the estimation target is a coast, the garbage related to the pollution level is, for example, drifting garbage. However, depending on the location of the estimation target, garbage other than drifting garbage may be the garbage related to the pollution level. The estimated pollution level is used for mechanical and quantitative monitoring of the location of the estimation target, and more specifically, for example, is used to determine the need for cleaning the coast.

[0020] Specifically, the contamination level estimation system 10 is configured with a server device, which is a computer including hardware such as a CPU (Central Processing Unit) and a memory. Each function of the contamination level estimation system 10, which will be described later, is realized by these components operating through programs or the like. The contamination level estimation system 10 may be realized by one computer, or may be realized by a computer system configured by connecting a plurality of computers to each other via a network.

[0021] Next, a description will be given of functions of the contamination level estimation system 10 according to this embodiment. As shown in FIG.

[0022] The image acquisition unit 11 is an image acquisition means for acquiring an image showing the location of the estimation target. The image is generated, for example, by a user who is an investigator capturing (taking a picture) with a camera at the location of the estimation target so that the location of the estimation target is captured. The image acquisition unit 11 acquires the image by receiving the image transmitted from the camera. The image acquisition unit 11 outputs the acquired image to the detection unit 12 and the contamination level estimation unit 13. The image acquisition unit 11 may standardize the acquired image to a preset size by resizing or the like for processing in the detection unit 12 and the contamination level estimation unit 13. The standardization may be performed using existing technology.

[0023] The capturing and transmission of the image may be performed by a smartphone used by the user using a dedicated application for estimating the pollution level (hereinafter referred to as a smartphone app). The camera does not need to be one provided in the smartphone, and may be an imaging device such as a consumer camera or a fixed camera installed so as to capture the location of the estimation target. The image acquisition unit 11 may also acquire images by a method other than the above.

[0024] The detection unit 12 is a detection means that detects dust parts (contaminated areas) in which dust is captured in an estimation target location in an image acquired by the image acquisition unit 11, and non-dust parts (non-contaminated areas) in which dust is not captured. The detection unit 12 may detect dust parts, non-dust parts, and parts other than the estimation target location by segmentation using machine learning. That is, the detection unit 12 may perform the above detection by artificial intelligence. The detection unit 12 performs detection, for example, as follows.

[0025] The detection unit 12 inputs an image from the image acquisition unit 11 and performs the above detection using a pre-stored trained model. The trained model is for segmentation (e.g., semantic segmentation) generated by machine learning. The trained model includes a neural network. The neural network may be multi-layered. That is, the neural network may be generated by deep learning. For example, the neural network is HRNet.

[0026] The trained model takes an image as input and outputs information indicating the classification of each pixel of the input image. The classification indicates what is captured in the pixel. The classification may be any classification that can distinguish between garbage, non-garbage, and areas other than the location of the estimation target. For example, there are eight classifications: man-made drifting garbage (man-made drifting objects), natural drifting garbage (natural drifting objects), sky, sea, sandy beach, existing natural objects (natural objects), existing man-made objects (installed objects), and other backgrounds. Of the above classifications, man-made drifting garbage and natural drifting garbage correspond to garbage areas. Of the above classifications, sandy beach, existing natural objects, and existing man-made objects correspond to non-garbage areas. Of the above classifications, sky, sea, and other backgrounds correspond to areas other than the location of the estimation target.

[0027] The trained model may be generated by an existing machine learning method. For example, a large number of (e.g., about 3,500) images of a beach containing washed-up trash are prepared as training images. The training images may be normalized as described above. In addition, data (pixel-unit area data) indicating which of the above classifications each pixel of the training images belongs to is generated as training data. Machine learning is performed using a data set that pairs the training images and the training data to generate a trained model.

[0028] The detection unit 12 inputs the image input from the image acquisition unit 11 into the trained model to obtain an output from the trained model. The output from the trained model is the detection result of dust and non-dust parts in the image. FIG. 2 shows two examples of the image input from the image acquisition unit 11 and the detection result by the detection unit 12. In the example shown in FIG. 2, a detection result 31 is obtained from an image 21, and a detection result 32 is obtained from an image 22. Note that the legend in FIG. 2 is for the detection results 31 and 32. The detection unit 12 outputs the obtained detection result to the contamination level estimation unit 13. Note that the detection unit 12 may detect dust and non-dust parts in the image by a method other than the above (for example, a pre-stored algorithm).

[0029] The contamination level estimation unit 13 is a contamination level estimation means that calculates the areas of dust parts and non-dust parts detected by the detection unit 12, and estimates the contamination level of the estimation target location based on the calculated areas. The contamination level estimation unit 13 may acquire information indicating the distance from the imaging point to each of the dust parts and non-dust parts in the image, and calculate the areas of the dust parts and non-dust parts weighted by the distances. The contamination level estimation unit 13 may calculate the distance from the image acquired by the image acquisition unit 11. The contamination level estimation unit 13 estimates the contamination level, for example, as follows.

[0030] The contamination level estimation unit 13 inputs an image from the image acquisition unit 11. The contamination level estimation unit 13 inputs a detection result from the detection unit 12. For each pixel of the image input from the image acquisition unit 11, the contamination level estimation unit 13 calculates the distance from the imaging point (the position of the imaging device at the time of imaging) to the part captured in the pixel, that is, the depth. The calculated depth does not need to be a value indicating the distance itself, but may be a value corresponding to the distance. The depth calculation may be performed only for pixels corresponding to at least the garbage part of the image (in the above example, the part of artificial drifting garbage and natural drifting garbage) and the non-garbage part (in the above example, the part of the sandy beach, existing natural objects, and existing artificial objects). The detection result from the detection unit 12 may be used to calculate the depth. The depth calculation may be performed by an existing method. The depth calculation may be performed by an approximate calculation using an approximate formula.

[0031] The detection unit 12 calculates the depth, for example, as follows. The detection unit 12 first determines the horizon in the image. The determination of the horizon may be performed by an existing method. For example, the determination of the horizon is performed based on the calculated amount of change in the up and down (vertical direction) of the average value of the horizontal axis (horizontal direction) of the boundary between the sea and the sky or other background, based on the detection result by the detection unit 12. Alternatively, a guide line predetermined in a smartphone app may be determined as the horizon. Next, the detection unit 12 calculates the depth of each pixel from the determined horizon by geometric calculation.

[0032] The contamination level estimation unit 13 may obtain information indicating the depth of each pixel of the image by a method other than the above. For example, the camera that captures the image may be equipped with a depth sensor, and the image obtained by the image acquisition unit 11 may be associated with information indicating the depth of each pixel, so that the contamination level estimation unit 13 may obtain the information.

[0033] Next, the pollution level estimation unit 13 calculates a depth-weighted area for each of the dust parts and non-dust parts detected by the detection unit 12. For example, the pollution level estimation unit 13 calculates the sum of the depth values ​​for each part as the depth-weighted area of ​​that part. That is, the pollution level estimation unit 13 calculates the sum of the depth values ​​for each pixel determined to be man-made drifting dust or natural drifting dust as the weighted area of ​​the dust part. The pollution level estimation unit 13 also calculates the sum of the depth values ​​for each pixel determined to be a sandy beach, an existing natural object, or an existing man-made object as the weighted area of ​​the non-dust part.

[0034] The pollution level estimation unit 13 estimates the pollution level of the estimation target location based on the calculated areas. For example, the pollution level estimation unit 13 calculates a ratio value between the weighted area of ​​the garbage part and the weighted area of ​​the non-garbage part. The pollution level estimation unit 13 calculates, for example, the ratio value (weighted area of ​​the garbage part) / (weighted area of ​​the non-garbage part) as the coverage area ratio of the drifting garbage. The larger this ratio value, the higher the pollution level is. The pollution level estimation unit 13 stores in advance a correspondence relationship between the ratio value and a plurality of levels indicating the pollution level, and estimates the pollution level from the ratio value based on the stored correspondence relationship. The above correspondence relationship may be a formula or conditional branch information for calculating the pollution level from the ratio value. Alternatively, the pollution level estimation unit 13 may use the calculated ratio value itself as the pollution level to be estimated.

[0035] The areas of the dust parts and non-dust parts do not have to be weighted by depth, for example, the number of pixels in the dust parts and non-dust parts may be used as the areas of the dust parts and non-dust parts.

[0036] Furthermore, the pollution level estimation unit 13 may calculate the number of pieces of garbage that appear in the location of the estimation target in the image, and estimate the pollution level of the location of the estimation target based on the calculated number of pieces of garbage. In this case, the pollution level estimation unit 13 calculates the number of pieces of garbage that appear in the image from the image input from the image acquisition unit 11. The garbage for which the number is to be calculated is the above-mentioned floating garbage. The calculation of the number of pieces of garbage may be performed by an existing method, for example, a technology for detecting objects from an image. Furthermore, the detection of the number of pieces of garbage may be performed together with the detection of the garbage portion by the above-mentioned detection unit 12 (in which case, the detection unit 12 will have a part of the function of the pollution level estimation means).

[0037] The contamination level estimation unit 13 estimates the contamination level based on the above area (or a value such as a ratio based on the area) and the calculated number of pieces of dust. For example, the contamination level estimation unit 13 inputs the above area value and the number of pieces of dust, stores in advance a calculation formula that outputs an estimated result of the contamination level, and estimates the contamination level using the calculation formula.

[0038] The above calculation formula (including the case where the number of pieces of dust is not used) may be a trained model generated by machine learning. That is, the pollution level estimation unit 13 may estimate the pollution level by artificial intelligence. The trained model may include a neural network. When using a trained model for estimating the pollution level, training data corresponding to the input (e.g., the value based on the area and the value of the number of pieces of dust) and output (information indicating the pollution level) of the trained model are prepared in advance, and machine learning is performed to generate the trained model. As the training data corresponding to the output, information indicating the pollution level estimated in advance by a system other than the pollution level estimation system 10 is used. Note that the pollution level estimation unit 13 may estimate the pollution level of the estimation target location based on the areas of the dust and non-dust parts by a method other than the above (e.g., a pre-stored algorithm).

[0039] Furthermore, the pollution level estimation unit 13 may estimate the pollution level of the estimation target location from each of a plurality of images, and estimate the pollution level of the estimation target location from the plurality of estimation results. For example, the pollution level estimation unit 13 may estimate the pollution level of the same coast from each of a plurality of images, and estimate the pollution level of the coast from the plurality of estimation results. In this case, the image acquired by the image acquisition unit 11 may be associated with information indicating the estimation target location, for example, information such as an ID that identifies the coast, and used in the above process. The estimation of the pollution level of the estimation target location from the plurality of estimation results may be performed by any method. For example, the plurality of estimation results may be averaged.

[0040] The contamination level estimation unit 13 outputs information indicating the contamination level of the estimation target location, which is the estimation result. For example, the contamination level estimation unit 13 transmits the estimation result to a smartphone that is the transmission source of the image used to estimate the contamination level. This transmission may be performed via a smartphone app. The contamination level estimation unit 13 may also transmit the detection result by the detection unit 12, for example, information in the form of an image such as the detection results 31 and 32 shown in FIG. 2. The user can grasp the contamination level of the estimation target location by referring to the transmitted information.

[0041] Furthermore, the pollution level estimation unit 13 may output the estimation results and the images (basis images) acquired by the image acquisition unit 11 and used for the estimation to a database accessible via a communication network such as the Internet. The estimation results may be made available for reference for each location of the estimation target, i.e., each coast. Furthermore, the pollution level estimation unit 13 may output to a method and destination other than those described above. These are the functions of the pollution level estimation system 10 according to this embodiment.

[0042] Next, a contamination level estimation method, which is a process executed by the contamination level estimation system 10 according to this embodiment (an operation method performed by the contamination level estimation system 10), will be described with reference to the flowchart of FIG.

[0043] In this process, the image acquisition unit 11 acquires an image showing a location to be estimated (S01, image acquisition step). Then, the detection unit 12 detects dust parts and non-dust parts in the image (S02, detection step). Then, the contamination level estimation unit 13 calculates the depth of each of the dust parts and non-dust parts in the image (S03, contamination level estimation step).

[0044] Next, the contamination level estimation unit 13 calculates the areas of the dust parts and non-dust parts weighted by depth (S04, contamination level estimation step). Next, the contamination level estimation unit 13 estimates the contamination level of the estimation target location based on the areas (S05, contamination level estimation step). Next, the contamination level estimation unit 13 outputs the estimation result of the contamination level of the estimation target location (S06). The above is the process executed by the contamination level estimation system 10 according to this embodiment.

[0045] In this embodiment, the contamination level of the estimation target location is estimated based on the areas of dust and non-dust parts detected from the image. This makes it possible to estimate the contamination level without relying on the subjectivity of the investigator. In other words, the contamination level can be estimated quantitatively and automatically based on a uniform standard, rather than on the investigator's empirical judgment. Therefore, this embodiment makes it possible to more accurately estimate the contamination level of the estimation target location.

[0046] Also, as in the above-described embodiment, an area weighted by depth may be calculated and used to estimate the contamination level. With this configuration, the areas of the dust and non-dust parts used to estimate the contamination level can be determined taking the depth into consideration. As a result, the contamination level of the target location can be estimated more accurately.

[0047] Also, as in the above-described embodiment, the depth may be calculated from the image used to estimate the contamination level. With this configuration, the weighted area can be calculated from the image alone. As a result, the contamination level of the location to be estimated can be easily and accurately estimated.

[0048] However, the depth does not have to be used when calculating the area. Even if the depth is used, the depth does not have to be calculated as described above, and information indicating the depth may be obtained by any method.

[0049] Also, as in the above-described embodiment, the number of dust particles in the image may be calculated and used to estimate the contamination level. With this configuration, the contamination level of the target location can be estimated more accurately. However, the number of dust particles does not have to be used to estimate the contamination level.

[0050] Furthermore, as in the above-described embodiment, dust parts, non-dust parts, and parts other than the location of the estimation target may be detected by segmentation using machine learning. With this configuration, the contamination level of the location of the estimation target can be reliably and appropriately estimated. However, the detection of dust parts and non-dust parts may be performed by a method other than the above.

[0051] Next, a description will be given of a contamination level estimation program for executing the above-mentioned series of processes by the contamination level estimation system 10. As shown in Fig. 4, the contamination level estimation program 100 is stored in a program storage area 111 formed in a computer-readable recording medium 110 that is inserted into a computer and accessed, or that is provided in the computer. The recording medium 110 may be a non-transitory recording medium.

[0052] The contamination level estimation program 100 includes an image acquisition module 101, a detection module 102, and a contamination level estimation module 103. Functions realized by executing the image acquisition module 101, the detection module 102, and the contamination level estimation module 103 are similar to the functions of the image acquisition unit 11, the detection unit 12, and the contamination level estimation unit 13 of the contamination level estimation system 10 described above, respectively.

[0053] The contamination level estimation program 100 may be configured such that a part or the whole of it is transmitted via a transmission medium such as a communication line and is received and recorded (including installed) by another device. Also, each module of the contamination level estimation program 100 may be installed in one of multiple computers, not just one computer. In that case, the above-mentioned series of processes are performed by a computer system consisting of the multiple computers.

[0054] The contamination level estimation system of the present disclosure has the following configuration. [1] A pollution level estimation system for estimating the pollution level of a target location, comprising: An image acquisition means for acquiring an image showing a location of an estimation target; a detection means for detecting a dust portion where dust is captured and a non-dust portion where dust is not captured in the image captured by the image capture means at a location to be estimated; a contamination level estimating means for calculating an area of ​​the dust portion and the non-dust portion detected by the detecting means, and estimating a contamination level of the estimation target location based on the calculated areas; A contamination level estimation system comprising: [2] The contamination level estimation system according to [1], wherein the contamination level estimation means obtains information indicating the distance from the imaging point to each of the dust parts and non-dust parts in the image, and calculates the areas of the dust parts and non-dust parts weighted by the distance. [3] The contamination level estimation system according to [2], wherein the contamination level estimation means calculates the distance from an image acquired by the image acquisition means. [4] The contamination level estimation system according to any one of [1] to [3], wherein the contamination level estimation means calculates the number of pieces of dust appearing in the location of the estimation target in the image, and estimates the contamination level of the location of the estimation target based on the calculated number of pieces of dust. [5] The contamination level estimation system according to any one of [1] to [4], wherein the detection means detects dust parts, non-dust parts, and parts other than the location to be estimated by segmentation using machine learning. [6] A pollution level estimation method which is an operating method of a pollution level estimation system for estimating a pollution level of a location to be estimated, comprising the steps of: An image acquisition step of acquiring an image showing the location of the estimation target; a detection step of detecting a dust portion where dust is captured and a non-dust portion where dust is not captured at a location to be estimated in the image acquired in the image acquisition step; a contamination level estimating step of calculating an area of ​​a non-dust portion of the dust portion detected in the detection step, and estimating a contamination level of the estimation target location based on the calculated area; A method for estimating the degree of contamination. [7] A contamination level estimation program for operating a computer as a contamination level estimation system for estimating the contamination level of a location to be estimated, comprising: The computer, An image acquisition means for acquiring an image showing a location of an estimation target; a detection means for detecting a dust portion where dust is captured and a non-dust portion where dust is not captured in the image captured by the image capture means at a location to be estimated; a contamination level estimating means for calculating an area of ​​the dust portion and the non-dust portion detected by the detecting means, and estimating a contamination level of the estimation target location based on the calculated areas; A contamination estimation program that functions as a. [Explanation of symbols]

[0055] 10...contamination level estimation system, 11...image acquisition unit, 12...detection unit, 13...contamination level estimation unit, 100...contamination level estimation program, 101...image acquisition module, 102...detection module, 103...contamination level estimation module, 110...recording medium, 111...program storage area.

Claims

1. A contamination level estimation system for estimating the contamination level of a location to be estimated, comprising: image acquisition means for acquiring an image showing the location to be estimated; detection means for detecting a garbage portion showing garbage at the location to be estimated and a non-garbage portion not showing garbage in the image acquired by the image acquisition means; contamination level estimation means for calculating the areas of the garbage portion and the non-garbage portion detected by the detection means and estimating the contamination level of the location to be estimated based on the calculated areas; and the contamination level estimation means acquires information indicating the distance from the imaging location to each of the garbage portion and the non-garbage portion in the image for each of the garbage portion and the non-garbage portion in the image, and calculates the areas of the garbage portion and the non-garbage portion weighted by the distance, a contamination level estimation system.

2. The contamination level estimation system according to claim 1, wherein the contamination level estimation means calculates the distance from the image acquired by the image acquisition means.

3. The contamination level estimation system according to claim 1 or 2, wherein the contamination level estimation means calculates the number of garbage items shown at the location to be estimated in the image and estimates the contamination level of the location to be estimated based also on the calculated number of garbage items.

4. The contamination level estimation system according to claim 1 or 2, wherein the detection means detects a garbage portion, a non-garbage portion, and a portion other than the location to be estimated by segmentation using machine learning.

5. A contamination level estimation method which is an operation method of a contamination level estimation system for estimating the contamination level of a location to be estimated, comprising: an image acquisition step of acquiring an image showing the location to be estimated; a detection step of detecting a garbage portion showing garbage at the location to be estimated and a non-garbage portion not showing garbage in the image acquired in the image acquisition step; a contamination level estimation step of calculating the areas of the garbage portion and the non-garbage portion detected in the detection step and estimating the contamination level of the location to be estimated based on the calculated areas; and in the contamination level estimation step, for each of the garbage portion and the non-garbage portion in the image, information indicating the distance from the imaging location to each of the garbage portion and the non-garbage portion is acquired, and the areas of the garbage portion and the non-garbage portion weighted by the distance are calculated, a contamination level estimation method.

6. A contamination level estimation program for operating a computer as a contamination level estimation system for estimating the contamination level of a location to be estimated, When the computer is made to function as image acquisition means for acquiring an image in which the location to be estimated is shown, detection means for detecting a garbage portion in which garbage at the location to be estimated in the image acquired by the image acquisition means is shown, and a non-garbage portion in which no garbage is shown, contamination degree estimation means for calculating the areas of the garbage portion and the non-garbage portion detected by the detection means and estimating the contamination degree of the location to be estimated based on the calculated areas, and the contamination degree estimation means is a contamination degree estimation program that acquires information indicating the distance from the imaging point to each of the garbage portion and the non-garbage portion in the image for each of the garbage portion and the non-garbage portion in the image, and calculates the areas of the garbage portion and the non-garbage portion weighted by the distance.

Images