JP2022106185A - Estimation device, estimation method and estimation program - Google Patents

Abstract

To easily estimate the mass of an insoluble substance from an image obtained by imaging fragments of the insoluble substance containing micro plastic.SOLUTION: An estimation device comprises: an acquisition unit which acquires a photographed image obtained by imaging an insoluble substance mixed in prescribed liquid; an object estimation unit which specifies an object that becomes the source of the insoluble substance and has flown out to the liquid on the basis of the appearance of the insoluble substance imaged in the photographed image; and a mass estimation unit which estimates the mass of the insoluble substance from the appearance of the insoluble substance acquired by the acquisition unit and the object estimated by the object estimation unit.SELECTED DRAWING: Figure 1

Description

The present invention relates to an estimation device, an estimation method and an estimation program.

In recent years, there are concerns about the environmental impact of so-called microplastics. Microplastics are microplastic pieces having a size of, for example, 5 to several mm or less, and are considered to be mainly caused by household wastewater and plastic waste flowing into rivers. As part of elucidating the actual situation of such microplastics and taking measures, microplastics are being collected from rivers. Specifically, for example, in order to measure the amount of microplastic per unit amount of water quickly and easily, an inspection device combining a net and a drainage meter is used.

Japanese Unexamined Patent Publication No. 2017-116260 Japanese Unexamined Patent Publication No. 2019-007901

Microplastics are discarded plastic waste and other small pieces that enter the river due to wind and rain and eventually flow into the sea. There are also types of microplastics, such as artificial turf and tire debris, that can be scraped and leaked even if they are not discarded. The effects of microplastics are said to be caused by predation by aquatic organisms such as fish. Furthermore, it has been pointed out that persistent organic pollutants attached to microplastics may adversely affect the ecosystem. Therefore, it is required to monitor the distribution and amount of microplastics over as wide a range as possible.

The present application has been made in view of the above, and an object thereof is to easily obtain the mass of an insoluble substance typified by microplastic.

The estimation device according to the present application is an object that is a source of an insoluble substance based on an acquisition unit that acquires a photographed image of an insoluble substance mixed in a predetermined liquid and an appearance of the insoluble substance photographed in the photographed image. It has an object estimation unit that identifies an object that has flowed out into a liquid, an appearance of an insoluble substance acquired by the acquisition unit, and a mass estimation unit that estimates the mass of the insoluble substance from the object estimated by the object estimation unit. It is characterized by.

According to one aspect of the embodiment, there is an effect that the mass of the insoluble substance can be easily estimated from the image obtained by capturing the fragments of the insoluble substance without using a high-precision mass measuring device or the like.

FIG. 1 is a diagram showing an example of estimation processing executed by the estimation device according to the embodiment. FIG. 2 is a diagram showing a configuration example of an estimation device according to an embodiment. FIG. 3 is a diagram showing an example of information registered in a database such as the appearance of the microplastic and the original object of the microplastic according to the embodiment. FIG. 4 is a diagram showing an example of information registered in a database such as an original object of microplastic according to an embodiment, image variations, and shape parameters. FIG. 5 is a diagram showing an example of the original object of the microplastic according to the embodiment, the image variation, and the information registered in the database of the mass calculation formula. FIG. 6 is a flowchart showing an example of the flow of processing executed by the estimation device according to the embodiment. FIG. 7 is a diagram showing an example of a hardware configuration.

Hereinafter, a mode for carrying out the estimation device, the estimation method, and the estimation program according to the present application (hereinafter referred to as “embodiment”) will be described in detail with reference to the drawings. It should be noted that this embodiment does not limit the estimation device, estimation method and estimation program according to the present application. Details of one or more embodiments are described in the following description and drawings. Further, each of the one or a plurality of embodiments can be appropriately combined as long as the processing contents do not contradict each other. Further, in the following one or a plurality of embodiments, the same parts are designated by the same reference numerals, and duplicate description will be omitted.

[1. About the estimation device]
First, an example of the processing executed by the estimation device 100 will be described with reference to FIG. FIG. 1 is a diagram showing an example of estimation processing executed by the estimation device according to the embodiment. In FIG. 1, microplastics contained in water are used as insoluble substances in flowing water such as rivers, canals, estuaries and seas, and in places where water exists such as lakes (hereinafter, may be collectively referred to as “collection points”). An example of a process of collecting and estimating the mass of the collected insoluble substance is described.

[1-1. Overview of estimation device]
The estimation device 100 is an information processing device capable of communicating with an arbitrary terminal device via a predetermined network such as the Internet, and is realized by, for example, a server device or a cloud system. The estimation device 100 may be able to communicate with any number of terminal devices via a network (for example, an Internet network).

The terminal device 200 is a terminal device used by a user, and may be any type of information processing device including, for example, a smartphone, a desktop PC (Personal Computer), a notebook PC, and a tablet PC. For example, the terminal device 200 has a function of receiving input of various information from a user, communicating with the estimation device 100 by wire or wirelessly, and outputting information as a communication result.

[1-2. About insoluble substances]
Here, the insoluble substance is a substance contained in the liquid containing the insoluble substance without being dissolved, and is, for example, a microplastic contained in water as a liquid. For example, the microplastic collected at the collection point leaks PET bottles dumped in rivers and the sea, and plastic products used in various facilities located near the river basin, and then deteriorates and collapses. It may occur by doing.

To give a more specific example, golf courses and tennis courts may be equipped with artificial turf made of plastic. However, when such artificial turf deteriorates due to use, a part of it may collapse and flow out to a river or the sea through an irrigation canal or the like when it rains. In paddy fields and fields, sustained-release fertilizer capsules in which fertilizer is coated with a plastic material such as polyethylene, capsules in which the pesticide base is uniformly coated with a polymer film, etc. (hereinafter, these pesticides, fertilizers, etc., are optional). The contents of the above coated with various plastic materials are simply collectively referred to as "capsules"). However, when it rains or drains water from a paddy field, such capsules may flow out into a river or the like and become microplastic.

In addition to the products mentioned above, plastic bags and plastic cups used in campgrounds, plastic trays for fishing tegs and foodstuffs, various illegally dumped plastic products, plastic waste dumped in the sea, etc. It may collapse and leak out as microplastic. In addition to these, there is a possibility that plastic products or a part of them may flow out into rivers or the like and collapse to become microplastics from facilities where various plastic products may be located.

In addition to these, there is a risk that industrial abrasives, abrasives, pellets, etc. will flow out as microplastics. In this way, what the outflowing object becomes microplastic as it is is called primary microplastic, and what the object of a certain plastic collapses into fine fragments is called secondary microplastic. In the following description, these primary microplastics and secondary microplastics are simply collectively referred to as microplastics, and no particular distinction is made.

[1-3. Collection of microplastics]
Microplastics are, for example, (1) those floating on the surface of the sea, (2) those floating in the sea, (3) those buried in sediments on the seabed, and (4) those mixed with sand on sandy beaches. , (5) Some marine organisms have entered the body. As a method for collecting microplastics, for example, (1) scoop the surface of seawater with a manta net or the like, (2) collect with a bongo net or the like, (3) collect with a hook scorer or the like, (4) a predetermined frame. There are methods such as collecting sand to a predetermined depth using, (5) dissecting an organism and removing the digestive tract.

Next, after performing predetermined pretreatments such as sieving, chemical treatment, and specific gravity separation, microscopic images and the like are acquired (step S10), and microplastics larger than about 100 μm are visually picked up one by one (step S10). Step S20). Those smaller than 100 μm cannot be picked up, so they are extracted using, for example, a micro-Fourier transform infrared spectrophotometer.

Here, the microplastic may be collected by a dedicated microplastic collecting device. For example, the collecting device has a structure in which a screw for flowing water backward and a net through which the water flowed by the screw passes are connected by a cylindrical duct. In addition, a water flow meter is installed behind the net to measure the amount of water that has passed through the net. By analyzing the microplastics that remain on the net when a certain amount of water passes through the net, the amount of microplastics located in ponds, rivers, etc. can be estimated.

[1-4. Measurement of microplastics]
The estimation device 100 inputs an image 30 of the microplastic (step S21), and measures the size (long side, short side, area, etc.) and shape (fibrous, debris, etc.) of each microplastic. Here, the image 30 _n (n is a natural number) is referred to as "image 30" unless an image is specified. For the microplastic photographed by using a digital camera attached to a microscope, the estimation device 100 measures the sizes of the long side, the short side, and the like using, for example, image processing software. Further, the estimation device 100 may measure the average particle size, area, etc. of the particles using image analysis software or the like to identify the shape (step S31).

[1-5. About object estimation processing]
When the applicant collected the microplastic and analyzed the collected microplastic, it was found that the object before collapse can be estimated to some extent based on the appearance such as its shape, color, size, and deterioration method. For example, microplastics produced by the collapse of artificial turf are often colored green. Further, when the microplastic is a capsule or a part thereof, a curved shape in a part or a dent when an object having a cavity collapses is often visually recognized.

In this way, the various appearances of the microplastic can be an index for estimating what kind of product the microplastic was (that is, what the original object was).

For example, the estimation device 100 estimates the object that is the source of the microplastic from the photographed image by using a model that has learned the relationship between the characteristics of the appearance of the microplastic and the object that is the source of the microplastic. (Step S32). For example, the estimation device 100 uses a model that learns the relationship between the characteristics of the color or shape of the microplastic and the object that is the source of the microplastic as the appearance of the microplastic, and uses a model that learns the relationship between the micro. Estimate the object that is the source of plastic.

By such an estimation process, the estimation device 100 can estimate the object that is the source of the microplastic, for example, the name of a product such as an artificial turf, a PET bottle, a plastic bag, or a capsule. As a result, when the object (product) is determined based on the microplastic, the estimation device 100 can acquire the material information from the original object information without performing the material analysis. On the other hand, when the original object is not determined, the estimation device 100 acquires the material information by using the dedicated device for material analysis.

[1-6. About material analysis of microplastics]
When it is not possible to visually determine the substance that is the source of the microplastic from the characteristics of the appearance of the microplastic, the estimation device 100 uses, for example, a micro-Fourier transform infrared spectrophotometer or the like to collect the microplastic. Is irradiated with infrared light, the amount of absorbed or reflected absorption is measured, molecular structure, functional group information and the like are acquired, and the material of the microplastic is analyzed (step S33).

When the estimation device 100 can determine the material of the microplastic, the estimation device 100 uses a model that learns the relationship between the characteristics and materials of the appearance of the microplastic and the object that is the source of the microplastic, and uses the model to be the source object. Is estimated (step S34).

[1-7. About mass estimation of microplastics]
The estimation device 100 calculates the volume of the microplastic by, for example, multiplying the measured area by the thickness of the microplastic. Further, the estimation device 100 calculates the mass of the microplastic by multiplying the volume by the estimated specific gravity of the original object.

To measure the thickness of microplastic, for example, it is measured using a microscope and a micrometer, but it is time-consuming and impractical. On the other hand, the size (long side, short side, area, etc.) and shape (fibrous, debris, etc.) of the microplastic can be measured by using, for example, image processing software. Therefore, for example, the estimation device 100 learns an empirical formula for calculating the mass using size and shape parameters, material information, and the like, and estimates the mass without using the measured value of the thickness of the microplastic. Calculate (step S35).

[1-8. About an example of various estimation processes]
Hereinafter, an example of the flow of the object estimation process and the mass estimation process executed by the estimation device 100 will be described with reference to FIG. In the following description, an example of performing object estimation processing and mass estimation processing on microplastic contained in water as a fluid having a flow such as a river or the sea will be described, but the embodiment is limited to this. It's not a thing. The object estimation process and mass estimation process described below are performed on insoluble substances contained in various predetermined liquids such as lakes and the like, microplastics contained in water having no flow (or having a slight flow). You may.

First, the user collects water flowing through a river or the like, and collects microplastics from the collected water. For example, the user collects microplatics with a pump, a water sampler, a plankton net, or the like.

The estimation device 100 measures the size such as color, long side, short side, and area of the microplastic image in the captured image 30, and identifies the shape such as fibrous shape and debris shape (step S31). Next, the estimation device 100 estimates the original object based on the size and shape of the microplastic (step S32). For example, the estimation device 100 may include characteristics of the shape of the photographed microplastic (for example, whether or not it has a curved or curved portion, whether or not it has a linear portion, etc.), microplastic. The color (for example, whether it is milky white or green), transparency, and the like are specified as the characteristics of the appearance of the microplastic (hereinafter, referred to as "appearance characteristics"). In addition, such a specific technique may be realized by various known image analysis techniques.

Subsequently, the estimation device 100 estimates the original object based on the specified appearance feature (step S32). For example, the estimation device 100 has a table in which an object that is the source of the microplastic and the appearance characteristics of the microplastic generated when the object flows out and collapses are associated in advance, and is similar to the specified appearance characteristics. The object associated with the appearance feature may be estimated as the original object of the photographed microplastic. Further, the estimation device 100 may use a model in which the relationship between the appearance of the microplastic and the original object is learned in advance. Such a model can be realized by, for example, a neural network trained to output information indicating the object (product) when an image of microplastic generated from a certain object (product) is input. Is.

The estimation device 100 may use the data of the material analysis of the microplastic. For example, the user analyzes the material of the microplastic using an infrared spectroscope or the like, and registers the material information indicating the material in the estimation device 100. In such a case, the estimation device 100 estimates the original object (product) based on the material (step S34). For example, when the material information indicating the material is input, the estimation device 100 outputs the information indicating the object that is the source of the microplastic when the information indicating the object that is the source of the microplastic is input. The estimation may be performed using the model trained as described above.

Subsequently, the estimation device 100 executes the mass estimation process of the microplastic (step S35). For example, a model formula for mass calculation is generated for each original object of microplastic. The parameters to be input to the mass calculation formula are mainly shape parameters indicating appearance information, material information, and shape mode of the microplastic. Appearance information is, for example, the measured area. The material information is the material information related to the material of the object when the original object (product) is known, and the material information measured by using an infrared spectroscope or the like when the original object is unknown. The shape mode is a shape parameter indicating a shape mode such as a fibrous shape or a fragment shape. Each mass calculation formula is a calculation formula for estimating the mass mainly based on the area, with the thickness of the microplastic for each object as a predetermined thickness. The thickness adopted by the calculation formula may be a thickness that varies depending on the area. The estimation device 100 estimates the mass of the estimated object using a calculation formula corresponding to the estimated object. In addition, depending on the object, for example, not only the area but also the material information such as polyethylene and polypropylene and the shape mode such as fibrous and debris affect the estimated mass even for the same object (product). give. The model formula related to the mass calculation in the embodiment of the present application can cope with such a situation. Such a calculation formula may be a model realized by, for example, a neural network or the like. In FIG. 1, for example, objects_01, object_02, and object_03 are illustrated, and for these, a model formula for mass calculation is generated for each object. In each calculation formula, the measured area value, material information, and shape mode are input as parameters, so even if the object is the same, for example, the estimated mass value corresponding to the material change due to deterioration and the shape mode change. Will be output.

In the above-described embodiment, the mass calculation formula for each object is used, but the calculation formula may be the mass calculation formula for each material which is the material. In the case of each material, the estimation device 100 estimates an object (product) from the appearance of the microplastic, identifies and registers the material of the estimated object. Then, the mass may be estimated from the area of the photographed microplastic using the mass calculation formula corresponding to the specified material. Further, the estimation device 100 registers the material of the microplastic analyzed by the user with an infrared spectroscopic device or the like. Then, the mass may be estimated from the area of the photographed microplastic by using the mass calculation formula corresponding to the registered material.

[1-9. Relationship between the deterioration level of objects and microplastics]
Further, the estimation device 100 may estimate the material and the original object (product) by using various information (that is, appearance features) obtained from the appearance of various microplastics. For example, the estimation device 100 may estimate the original object based not only on the shape, color, and material, but also on the deterioration level and size of the microplastic. For example, the estimation device 100 may use a "capsule" when the microplastic is a milky white spherical shell-like object or a part thereof, and the original sphere is estimated to be about 2 mm to 3 mm in size. It may be an estimated object.

Further, in the estimation device 100, when the microplastic is a light blue sheet-like substance and the thickness is within the range of about 0.08 mm to 0.3 mm, the estimation device 100 is used for home repair, camping, etc. A "sheet" may be used as an estimated object. Further, in the estimation device 100, when the microplastic has a white pellet-like shape and the component is polyethylene, the estimation device 100 may use the “raw material pellet” which is the raw material of the plastic product as the estimation object. Such "raw material pellets" are often stored outdoors in factories that manufacture plastic products. Therefore, the estimation device 100 may presume that the "raw material pellet" is a microplastic leaked from a plastic product manufacturing factory.

Further, in the estimation device 100, when the microplastic is white polyurethane and the thickness is about 2 mm, a "plastic cup" may be used as an estimation object. Further, in the estimation device 100, when the microplastic is polyethylene in the form of a string and the diameter of the transition constituting the microplastic is about 0.3 mm, the polyethylene rope or the transition constituting the polyethylene rope may be used as the estimation object. good.

In addition, it is considered that the shape and transparency of various plastic products change as a result of partial collapse or deterioration as they are carried by a stream of water. It can be considered that such various deterioration modes differ depending on the composition, shape, and the like of the plastic product. Therefore, the estimation device 100 may estimate the original object by using the deterioration mode of the microplastic as the appearance of the microplastic. For example, the estimation device 100 may estimate the object that is the source of the microplastic from the photographed image by using the model that has learned the relationship between the deterioration mode and the object that is the source of the microplastic.

For example, the estimation device 100 learns a model for estimating the original object when the degree of collapse set by comparing the original object and the captured image of the microplastic is input together with the image of the microplastic. .. Such a degree of collapse is set manually, for example. Then, the estimation device 100 may estimate the original object by inputting the captured image input by the user and the degree of collapse into the model.

Further, the estimation device 100 may estimate the object that is the source of the microplastic based on the collection point where the microplastic is collected. For example, when there are many golf courses and tennis courts around the river, there is a high possibility that microplastics based on artificial turf will be collected in the lower reaches of the river rather than capsules. On the other hand, when there are many paddy fields around the river, there is a high possibility that microplastics based on capsules will be collected downstream of the river.

Further, the estimation device 10 may estimate the object that is the source of the microplastic from the material information indicating the component of the microplastic. For example, the estimation device 10 stores components in advance for each candidate of the estimated object, and is based on an external image from among the candidates whose similarity with the collected components of the microplastic exceeds a predetermined threshold value. , The estimated object may be specified.

In addition to these information, the estimation device 100 estimates the original object of the microplastic based on various arbitrary information if the original object is estimated in consideration of at least the appearance of the microplastic. You can do it. Further, the estimation device 100 may use various arbitrary objects such as a plastic net and a plastic cup as the estimation object, and may use polyethylene, polypropylene, and an object of any material as the estimation object.

[2. Configuration of estimation device]
Next, a configuration example of the estimation device 100 according to the embodiment will be described with reference to FIG. FIG. 2 is a diagram showing a configuration example of the estimation device 100 according to the embodiment. As shown in FIG. 2, the estimation device 100 includes a communication unit 210, a storage unit 220, and a control unit 230. The estimation device 100 has an input unit (for example, a keyboard, a mouse, etc.) that receives various operations from an administrator or the like who uses the estimation device 100, and a display unit (liquid crystal display, etc.) for displaying various information. You may.

(Communication unit 210)
The communication unit 210 is realized by, for example, a NIC (Network Interface Card) or the like. The communication unit 210 is connected to the network network by wire or wirelessly, and transmits / receives information to / from the terminal device 200 via the network network.

(Memory unit 220)
The storage unit 220 is realized by, for example, a semiconductor memory element such as a RAM (Random Access Memory) or a flash memory (Flash Memory), or a storage device such as a hard disk or an optical disk. As shown in FIG. 2, the storage unit 220 has a microplastic appearance / object database 221, an object estimation model 222, a mass calculation model 223, and a mass calculation formula database 224.

(Appearance of microplastic / object database 221)
In the microplastic appearance / object database 221, for example, as shown in FIG. 3, as the appearance of the microplastic, the image "image_01", the color information "color_01", the long side "l_max_01", the short side "l_min_01", and the area " Area_01 ", shape" shape_01 ", material" material_01 "as material information, candidate 1" object_011 ", candidate 2" object_012 ", candidate 3" object_013 "as candidates for the original object (product name), etc. are associated with each other. It is remembered. For example, if the original object of the collected microplastic can be identified only by visual inspection, the fact that the original object is visually identified is stored in Candidate 1 of the original object.

(Object estimation model 222)
The object estimation model 222 is a model that estimates the original object from the captured image of the microplastic. The original object is estimated from the characteristics of the microplastic appearance captured in the captured image. For example, the material information is taken into consideration in the appearance information such as the color information, the long side, the short side, the area, and the shape of the microplastic, and the candidate of the original object (product name) is output. An existing object estimation model may be used. When targeting microplastics, for example, a model that can capture information that can be visually determined may be used as appropriate. For example, as shown in FIG. 4, the object “object_01”, the image variations “image_01” “a” to “d”, and the shape parameters “shape_a_01” to “shape_d_01” are stored. The shape parameter is a parameter indicating a shape mode such as a fibrous shape or a fragment shape. For example, in the case of the object “object_01”, the image variations “image_01” “a” to “d” correspond to the shape parameters “shape_a_01” to “shape_d_01”, respectively. By calculating the similarity between the captured image and the image variation, it is possible to narrow down the candidates for the estimated object having a high degree of similarity. Further, since the estimation device 100 can select the most similar image from the image variations for the input image and at the same time obtain the shape parameter corresponding to the image, the mass estimation related to the microplastic described later is performed. It is also possible to obtain the parameters required for the calculation formula.

(Mass calculation model 223)
In the mass calculation model 223, the learning unit 234 learns the relationship between the parameters regarding the appearance information, material information, shape mode, etc. of the microplastic, which are the parameters necessary for generating the mass calculation formula, and the mass calculation formula database is used. offer. Further, by using the data registered in the appearance / object database 221 of the microplastic, the parameter value and the calculation formula model suitable for the mass calculation formula are provided each time the material information or the shape mode changes. In addition, visual information and empirical values are also incorporated to provide appropriate parameter values and calculation formula models for mass calculation formulas.

(Mass calculation formula database 224)
In the mass calculation formula database 224, for example, as shown in FIG. 5, an object "object_01", an image variation "image_01""a" to "d", a mass estimation calculation formula "function_01", and the like are stored in association with each other. ing. In the mass calculation formula "function_01", the area, material information, shape parameters, etc. are input, and the estimated value of the mass is output. For example, using the mass calculation formula database 224, the calculation formula is specified based on the original object estimated from the image and the shape mode of the microplastic. The measured value of the area is input to the calculation formula, and the mass of the microplastic is output. In the mass calculation formula database 224, a mass calculation formula for each object is used, but the calculation formula may be a mass calculation formula for each material which is a material.

For the shape aspect, the shape parameter specified by the mass calculation model 223 is used. Specifically, in the mass calculation model 223, a shape parameter corresponding to an image having a shape most similar to the shape of the target microplastic is selected from a plurality of images of microplastics associated with the original object. ..

(Control unit 230)
The control unit 230 is a controller, and is stored in a storage device inside the estimation device 100 by a processor such as a CPU (Central Processing Unit) or an MPU (Micro Processing Unit). (Corresponding to one example) is realized by executing RAM or the like as a work area. Further, the control unit 230 is a controller, and may be realized by, for example, an integrated circuit such as an ASIC (Application Specific Integrated Circuit) or an FPGA (Field Programmable Gate Array).

As shown in FIG. 2, the control unit 230 has an acquisition unit 231, an object estimation unit 232, a mass estimation unit 233, and a learning unit 234, and realizes or executes the functions and actions of information processing described below. do. The internal configuration of the control unit 230 is not limited to the configuration shown in FIG. 2, and may be any other configuration as long as it is configured to perform information processing described later.

(Acquisition unit 231)
The acquisition unit 231 acquires a microscope image as image data using, for example, a microscope. For example, the acquisition unit 231 acquires the result of the material analysis of the microplastic. In addition, the acquisition unit 231 acquires information on objects used in factories and the like, object information that may be the source of microplastics collected by the user from the terminal device 200 used by the user, and the like. It may be registered in the appearance / object database 221 of the microplastic.

(Object estimation unit 232)
The object estimation unit 232 estimates the object that is the source of the microplastic and that has flowed out into the water, based on the appearance of the microplastic taken in the microscope image. For example, the object estimation unit 232 uses a model that learns the relationship between the color information, area, shape, etc. of the appearance of the microplastic and the object that is the source of the microplastic, and uses a model that learns the relationship between the microplastic image and the microplastic. Estimate the original object. For example, the object estimation unit 232 inputs the microscope image acquired by the acquisition unit 231 into the object estimation model 222 trained by the learning unit 234, so that the object estimation model 222 estimates the object from the appearance of the microplastic. Get information about.

The object estimation unit 232 may estimate the original object from the characteristics of the appearance of the microplastic without using the object estimation model 222. For example, the object estimation unit 232 identifies appearance features showing various appearance features such as shape, color, size, thickness, transparency, etc. of the microplastic by image analysis. Then, the object estimation unit 232 may use an object whose appearance feature is similar to that of the specified appearance feature among the objects registered in advance as the estimation object.

Further, the object estimation unit 232 may estimate the object that is the source of the microplastic based on the appearance of the microplastic and the collection point. For example, the object estimation unit 232 identifies a factory within a predetermined range from the collection point, a factory located upstream of the collection point, and the like. Subsequently, the object estimation unit 232 identifies an object associated with the specified factory based on the use and sales history of the object as a candidate object. Then, the object estimation unit 232 may use an object corresponding to the appearance of the microplastic as an estimation object from the candidate objects. Further, the object estimation unit 232 may have the object estimation model 222 calculate the possibility that each object is the source of the microplastic, and among the candidate objects, the one with the highest possibility may be used as the estimation object.

(Mass estimation unit 233)
The mass estimation unit 233 estimates the mass of a single microplastic from the area of the microplastic, material information, shape parameters, and the like. Basically, the thickness of the microplastic is measured in advance, the area of the collected microplastic is multiplied by the thickness to calculate the volume, and the density is obtained from the material information and multiplied by the volume. The mass can be calculated by this. However, it is troublesome to measure the thickness of microplastics using a microscope and a micrometer every time they are collected. In the case of microplastics, it is empirically known that it is possible to construct a model calculation formula that calculates mass by inputting area, material information, shape parameters, etc., so using the model calculation formula registered in the database, micro Estimate the mass of plastic.

The area used in the calculation formula can be easily obtained by, for example, existing image processing software. As for the material information, if the original object is known, the material information can be used, and even if the original object is not known, it can be acquired by using a micro-Fourier transform infrared spectrophotometer or the like. Therefore, for example, if the shape parameters and the like are properly learned, the mass with constant accuracy can be calculated by the model calculation formula.

(Learning Department 234)
The learning unit 234 learns the object estimation model 222. For example, the learning unit 234 learns the object estimation model 222 so as to output information indicating an object that is the source of the microplastic estimated by hand or the like when a microscope image of the microplastic is input. The learning unit 234 specifies, for example, appearance information such as color information, area, and shape of microplastic from a microscope image, and when the appearance information is input, the learning unit 234 outputs information indicating the original object. The estimation model 222 may be trained. When the object estimation model 222 obtained by such learning is used, the object estimation unit 232 generates appearance information from the microscope image and inputs the generated appearance information to the object estimation model 222 to input the generated appearance information to the original object. Will be estimated.

Further, the learning unit 234 may train the mass calculation model 223 to learn the relationship between the shape of the microplastic and the object that is the source of the microplastic as the shape of the microplastic. The shape aspect is, for example, fibrous, debris, or the like. For example, the learning unit 234 specifies the shape aspect of the microplastic based on the analysis result of the microscope image, and when the specified shape aspect and the microscope image are input, the mass calculation is performed so as to output the mass of the microplastic. The model 223 may be trained. When the mass calculation model 223 in which such learning has been performed is used, the mass estimation unit 233 identifies the shape mode from the microscope image and inputs the specified shape mode to the mass calculation model 223 to input the mass of the microplastic. Will be estimated.

[3. Estimating processing flow]
Next, the procedure of the estimation process by the estimation device 100 according to the embodiment will be described.
FIG. 6 is a flowchart showing a processing procedure in which the estimation device 100 according to the embodiment inputs an image of the microplastic, estimates the original object, and estimates the mass of the microplastic.

First, the acquisition unit 231 of the estimation device 100 acquires an image of the microplastic (step S101). Specifically, a single piece of microplastic having a size of about 100 μm is picked up from the microscope image, and an image of the single piece of microplastic is obtained.

Next, the acquisition unit 231 of the estimation device 100 acquires color information of the microplastic, size information such as a long side, a short side, and an area, and identification information of a shape such as a fibrous shape and a fragment shape (step S102). Specifically, for example, measurement and identification are performed using image analysis software or the like.

Next, the acquisition unit 231 of the estimation device 100 acquires the material analysis result of the microplastic (step S103). If the object that is the source of the microplastic is known and the material information is already obtained, the material analysis is not performed.

Next, the object estimation unit 232 of the estimation device 100 estimates the object that is the source of the microplastic (step S104). Specifically, the object estimation unit 232 estimates while learning by the learning unit 234 in consideration of not only the color, shape, and material but also the deterioration and size of the microplastic.

Next, the mass estimation unit 233 of the estimation device 100 estimates the mass of the microplastic (step S105). Specifically, the area of the microplastic, the material information of polyethylene, polypropylene, etc., and the shape parameters such as fibrous and debris are input to the mass calculation formula learned by the learning unit 234 to estimate the mass.

Next, the estimation device 100 provides the mass estimated by the mass estimation unit 233 to the terminal device 200 (step S106).

[4. others〕
Further, among the processes described in the above-described embodiment, a part of the processes described as being automatically performed can also be performed manually. Alternatively, all or part of the process described as being performed manually can be automatically performed by a known method. In addition, information including processing procedures, specific names, various data and parameters shown in the above documents and drawings can be arbitrarily changed unless otherwise specified. For example, the various information shown in each figure is not limited to the information shown in the figure.

Further, each component of each of the illustrated devices is a functional concept, and does not necessarily have to be physically configured as shown in the figure. That is, the specific form of distribution / integration of each device is not limited to the one shown in the figure, and all or part of them may be functionally or physically distributed / physically in any unit according to various loads and usage conditions. Can be integrated and configured.

For example, a part or all of the storage unit 220 shown in FIG. 2 may not be held by the estimation device 100, but may be held by a storage server or the like. In this case, the estimation device 100 acquires various information such as traveling information by accessing the storage server.

[5. Hardware configuration]
Further, the estimation device 100 according to the above-described embodiment is realized by, for example, a computer 1000 having a configuration as shown in FIG. 7. FIG. 7 is a diagram showing an example of a hardware configuration. The computer 1000 is connected to the output device 1010 and the input device 1020, and the arithmetic unit 1030, the primary storage device 1040, the secondary storage device 1050, the output IF (Interface) 1060, the input IF 1070, and the network IF 1080 are connected by the bus 1090. Have.

The arithmetic unit 1030 operates based on a program stored in the primary storage device 1040 or the secondary storage device 1050, a program read from the input device 1020, or the like, and executes various processes. The primary storage device 1040 is a memory device such as a RAM that temporarily stores data used by the arithmetic unit 1030 for various operations. Further, the secondary storage device 1050 is a storage device in which data used by the calculation device 1030 for various calculations and various databases are registered, and is realized by a ROM (Read Only Memory), an HDD, a flash memory, or the like.

The output IF 1060 is an interface for transmitting information to be output to an output device 1010 that outputs various information such as a monitor and a printer. For example, USB (Universal Serial Bus), DVI (Digital Visual Interface), and the like. It is realized by a connector of a standard such as HDMI (registered trademark) (High Definition Multimedia Interface). Further, the input IF 1070 is an interface for receiving information from various input devices 1020 such as a mouse, a keyboard, a scanner, and the like, and is realized by, for example, USB.

The input device 1020 is, for example, an optical recording medium such as a CD (Compact Disc), a DVD (Digital Versatile Disc), a PD (Phase change rewritable Disk), a magneto-optical recording medium such as an MO (Magneto-Optical disk), or a tape. It may be a device that reads information from a medium, a magnetic recording medium, a semiconductor memory, or the like. Further, the input device 1020 may be an external storage medium such as a USB memory.

The network IF 1080 receives data from another device via the network N and sends it to the arithmetic unit 1030, and also transmits the data generated by the arithmetic unit 1030 to the other device via the network N.

The arithmetic unit 1030 controls the output device 1010 and the input device 1020 via the output IF 1060 and the input IF 1070. For example, the arithmetic unit 1030 loads a program from the input device 1020 or the secondary storage device 1050 onto the primary storage device 1040, and executes the loaded program.

For example, when the computer 1000 functions as the estimation device 100, the arithmetic unit 1030 of the computer 1000 realizes the function of the control unit 230 by executing the program loaded on the primary storage device 1040.

[6. effect〕
As described above, the estimation device 100 according to the embodiment includes an acquisition unit 231, an object estimation unit 232, a mass estimation unit 233, and a learning unit 234.

In the estimation device 100 according to the embodiment, the acquisition unit 231 acquires a photographed image of the insoluble substance mixed in the predetermined liquid, and the object estimation unit 232 captures the appearance of the insoluble substance photographed in the photographed image. Based on this, an object that is the source of the insoluble substance and that has flowed out into the liquid is identified, and the mass estimation unit 233 uses the appearance of the insoluble substance acquired by the acquisition unit 231 and the object estimated by the object estimation unit 232. Estimate the mass of insoluble material.

As a result, the acquisition unit 231 acquires a photographed image of the insoluble substance mixed in the predetermined liquid, and the object estimation unit 232 acquires the photographed image of the insoluble substance based on the appearance of the insoluble substance photographed in the photographed image. The original object that has flowed out into the liquid is identified, and the mass estimation unit 233 estimates the appearance of the insoluble substance acquired by the acquisition unit 231 and the mass of the insoluble substance from the object estimated by the object estimation unit 232. Therefore, the mass of the insoluble substance can be easily estimated from the image obtained by capturing the fragments of the insoluble substance without using a high-precision mass measuring device or the like.

Further, in the estimation device 100 according to the embodiment, the mass estimation unit 233 estimates the thickness of the insoluble substance based on the object specified by the object estimation unit 232, and determines the estimated thickness and the appearance of the insoluble substance. Based on this, the mass of the insoluble substance is estimated.

As a result, the mass estimation unit 233 estimates the thickness of the insoluble substance based on the object specified by the object estimation unit 232, and determines the mass of the insoluble substance based on the estimated thickness and the appearance of the insoluble substance. For the estimation, the mass of the microplastic can be easily estimated from the image obtained by capturing the fragments of the insoluble substance without using a high-precision mass measuring device or the like.

Further, in the estimation device 100 according to the embodiment, the acquisition unit 231 acquires a photographed image of the microplastic mixed in the predetermined liquid as an insoluble substance, and the object estimation unit 232 is based on the appearance of the microplastic. , The object that is the source of the microplastic is specified, and the mass estimation unit 233 estimates the mass of the microplastic.

As a result, the acquisition unit 231 acquires a photographed image of the microplastic mixed in the predetermined liquid as an insoluble substance, and the object estimation unit 232 acquires the object that is the source of the microplastic based on the appearance of the microplastic. Since the mass estimation unit 233 estimates the mass of the microplastic, the mass of the microplastic can be easily estimated from the image obtained by capturing the fragments of the insoluble substance without using a high-precision mass measuring device or the like. ..

Further, in the estimation device 100 according to the embodiment, the acquisition unit 231 further acquires the material information of the insoluble substance, and the object estimation unit 232 becomes the source of the insoluble substance based on the material information acquired by the acquisition unit 231. An object that has flowed out into a liquid is estimated, and the mass estimation unit 233 estimates the mass of an insoluble substance from the material information acquired by the acquisition unit 231 and the object estimated by the object estimation unit 232.

As a result, the acquisition unit 231 further acquires the material information of the insoluble substance, and the object estimation unit 232 is an object that is the source of the insoluble substance and flows out to the liquid based on the material information acquired by the acquisition unit 231. The mass estimation unit 233 estimates the mass of the insoluble substance from the material information acquired by the acquisition unit 231 and the object estimated by the object estimation unit 232, so that the mass estimation unit 233 uses a high-precision mass measuring device or the like. The mass of the insoluble substance can be easily estimated from the image of the fragment of the insoluble substance.

Further, in the estimation device 100 according to the embodiment, the object estimation unit 232 estimates the material of the insoluble substance based on the appearance of the insoluble substance acquired by the acquisition unit 231, and the liquid is based on the estimated material and appearance. Estimate the leaked object.

As a result, the object estimation unit 232 estimates the material of the insoluble substance based on the appearance of the insoluble substance acquired by the acquisition unit 231, and estimates the object that has flowed out into the liquid based on the estimated material and appearance. , The original object of the insoluble substance can be easily estimated.

Further, in the estimation device 100 according to the embodiment, the object estimation unit 232 uses a model that learns the relationship between the size and material of the insoluble substance and the object that is the source of the insoluble substance, and is insoluble from the photographed image. Estimate the object that is the source of the substance.

As a result, the object estimation unit 232 estimates the object that is the source of the insoluble substance from the captured image using the model that learned the relationship between the size and material of the insoluble substance and the object that is the source of the insoluble substance. Therefore, the original object of the insoluble substance can be easily estimated.

Further, in the estimation device 100 according to the embodiment, the mass estimation unit 233 estimates the mass of the insoluble substance based on the area of the insoluble substance in the captured image and the object that is the source of the insoluble substance specified by the object estimation unit. do.

As a result, the mass estimation unit 233 estimates the mass of the insoluble substance based on the area of the insoluble substance in the captured image and the object that is the source of the insoluble substance specified by the object estimation unit 232, so that the mass is highly accurate. The mass of the insoluble substance can be easily estimated from the image of the fragment of the insoluble substance without using a measuring instrument or the like.

Further, in the estimation device 100 according to the embodiment, the mass estimation unit 233 is a model for estimating the mass of the insoluble substance from the appearance of the insoluble substance, and is among a plurality of models different for each object that is the source of the insoluble substance. Using the model corresponding to the object estimated by the object estimation unit 232, the mass of the insoluble substance is estimated from the appearance of the insoluble substance in the photographed image acquired by the acquisition unit 231.

As a result, the mass estimation unit 233 is a model for estimating the mass of the insoluble substance from the appearance of the insoluble substance, and is estimated by the object estimation unit 232 among a plurality of models different for each object that is the source of the insoluble substance. Fragments of insoluble substances are estimated from the appearance of the insoluble substances in the captured image acquired by the acquisition unit 231 using a model corresponding to the object, without using a high-precision mass measuring instrument or the like. The mass of the insoluble substance can be easily estimated from the image taken.

By each of the above-mentioned processes, the estimation device 100 acquires a photographed image of the insoluble substance mixed in the predetermined liquid, and becomes the source of the insoluble substance based on the appearance of the insoluble substance photographed in the photographed image. It has the effect of identifying an object that has flowed out into a liquid, and estimating the appearance of the insoluble substance and the mass of the insoluble substance from the identified object.

Although some of the embodiments of the present application have been described in detail with reference to the drawings, these are examples, and various modifications are made based on the knowledge of those skilled in the art, including the embodiments described in the disclosure column of the invention. It is possible to carry out the present invention in other modified forms.

Further, the estimation device 100 described above may be realized by a plurality of server computers, and depending on the function, an external platform or the like may be called by API (Application Programming Interface), network computing, or the like to realize the configuration. It can be changed flexibly.

Further, the above-mentioned "section, module, unit" can be read as "means" or "circuit". For example, the acquisition unit can be read as an acquisition means or an acquisition circuit.

1 Estimating system 100 Estimating device 200 Terminal device 210 Communication unit 220 Storage unit 221 Microplastic appearance / object database 222 Object estimation model 223 Mass calculation model 224 Mass calculation formula database 230 Control unit 231 Acquisition unit 232 Object estimation unit 233 Mass estimation unit 234 Learning Department

Claims (10)

An acquisition unit that acquires a photographed image of an insoluble substance that was mixed in a predetermined liquid, and
An object estimation unit that identifies an object that is the source of the insoluble substance and that has flowed out into the liquid, based on the appearance of the insoluble substance captured in the captured image.
An estimation device comprising the appearance of the insoluble substance acquired by the acquisition unit and a mass estimation unit that estimates the mass of the insoluble substance from the object estimated by the object estimation unit. The mass estimation unit estimates the thickness of the insoluble substance based on the object specified by the object estimation unit, and determines the mass of the insoluble substance based on the estimated thickness and the appearance of the insoluble substance. The estimation device according to claim 1, wherein the estimation is performed. The acquisition unit acquires a photographed image of microplastic mixed in a predetermined liquid as the insoluble substance.
The object estimation unit identifies the object that is the source of the microplastic based on the appearance of the microplastic.
The estimation device according to claim 1 or 2, wherein the mass estimation unit estimates the mass of the microplastic. The acquisition unit further acquires the material information of the insoluble substance, and obtains the material information.
Based on the material information acquired by the acquisition unit, the object estimation unit estimates an object that is a source of the insoluble substance and has flowed out into the liquid.
One of claims 1 to 3, wherein the mass estimation unit estimates the mass of the insoluble substance from the material information acquired by the acquisition unit and the object estimated by the object estimation unit. The estimation device described in 1. The object estimation unit estimates the material of the insoluble substance based on the appearance of the insoluble substance acquired by the acquisition unit, and estimates the object that has flowed out into the liquid based on the estimated material and the appearance. The estimation device according to any one of claims 1 to 3, wherein the estimation device is used. The object estimation unit estimates the object that is the source of the insoluble substance from the photographed image by using a model that has learned the relationship between the appearance and material of the insoluble substance and the object that is the source of the insoluble substance. The estimation device according to claim 5, wherein the estimation device is used. The mass estimation unit is characterized in that the mass of the insoluble substance is estimated based on the area of the insoluble substance in the captured image and the object that is the source of the insoluble substance specified by the object estimation unit. The estimation device according to any one of claims 1 to 6. The mass estimation unit is a model for estimating the mass of the insoluble substance from the appearance of the insoluble substance, and is specified by the object estimation unit among a plurality of models different for each object that is the source of the insoluble substance. One of claims 1 to 7, wherein the mass of the insoluble substance is estimated from the appearance of the insoluble substance in the photographed image acquired by the acquisition unit using a model corresponding to the object. The estimation device described. An estimation method performed by a computer
The acquisition process to acquire a photographed image of an insoluble substance mixed in a predetermined liquid, and
An object estimation step for identifying an object that is the source of the insoluble substance and that has flowed out into the liquid based on the appearance of the insoluble substance captured in the captured image.
An estimation method comprising: the appearance of the insoluble substance acquired by the acquisition step and a mass estimation step of estimating the mass of the insoluble substance from the object estimated by the object estimation step. The acquisition procedure to acquire a photographed image of an insoluble substance mixed in a predetermined liquid, and
An object estimation procedure for identifying an object that is the source of the insoluble substance and that has flowed out into the liquid based on the appearance of the insoluble substance captured in the captured image.
An estimation program comprising causing a computer to execute the appearance of the insoluble substance acquired by the acquisition procedure and the mass estimation procedure of estimating the mass of the insoluble substance from the object estimated by the object estimation procedure.

Images