JP2019204168A - Sorting device, sorting method, sorting program, and computer readable recording medium - Google Patents

Abstract

To allow users to easily set and change sorting accuracy.SOLUTION: A sorting device comprises a learning data creating unit 122, a learning unit 123, a sorting-target selection unit 124, a determination unit 125, and an air injecting nozzle 14, where the learning data creating unit creates the learning data LD from a line sensor camera 11 and from classified object data obtained by the line sensor camera 11, the learning unit sorts the mixture by type by using the learning data LD, learns a method to identify by type and creates a learning model GM digitalizing knowledge and experience obtained by the learning, the sorting-target selection unit to sort the type of the sorting-target, the determination unit calculates the recognition rate RR from the mixture data obtained by the line sensor camera 11 on the basis of the learning model GM and determines the existence and the position of the sorting-target on the basis of the recognition rate RR, and the air injecting nozzle sorts the sorting-target from the mixture MO on the basis of a threshold provided for the determination result and the recognition rate RR of the determination unit 125.SELECTED DRAWING: Figure 8

Description

  The present invention relates to a sorting device, a sorting method, a sorting program, and a computer-readable recording medium that sort a sorting target object from a mixture composed of a plurality of types of objects.

  In recent years, recycling that recycles wastes and uses them as raw materials for new products has been carried out by many companies for the purpose of environmental protection and improvement of corporate image.

  Although the recycling field is diverse, for example, in the field of recycling recycled paper to produce recycled paper, there is a problem of impurities such as, for example, when plastic such as laminate is mixed in the used paper, the purity of the paper is lowered. In addition, when harmful substances are mixed, the harmful substances are diffused widely. For this reason, before recycling, a step of selecting an object and impurities used as raw materials is required. Further, it is required that a sorting object can be freely sorted according to a recycling application, for example, sorting white paper and colored paper.

  In addition, regardless of recycling, it is necessary to select good products and defective products at the time of product manufacture. Therefore, it can be said that the technology of selecting an object into two or more is one of the indispensable technologies in the manufacturing industry. Techniques for selecting this type of object into two or more types are disclosed in Patent Document 1 and Patent Document 2, for example.

  Patent Document 1 discloses a technique related to a sorting apparatus that includes a detection unit including a light source and an optical sensor and sorts an object based on the luminance of reflected light.

  Patent Document 2 discloses a technology related to a gravity sensor and a sorting device that includes an RGB camera, an X-ray camera, a near-infrared camera, and a 3D camera as an imaging device, and automatically sorts an object using artificial intelligence.

JP 2018-017639 A JP 2017-109197 A

  However, the sorting device disclosed in Patent Document 1 needs to set in advance a standard and algorithm for sorting objects based on the brightness of reflected light, and these settings require specialized knowledge and experience. For this reason, the user cannot easily set or change the setting.

  In addition, the sorting device disclosed in Patent Document 2 using artificial intelligence does not require the setting as described above, but requires a step of learning a criterion and a method for sorting into artificial intelligence in advance, so that the user can easily It was not an aspect that could be set.

  As described above, in the conventional sorting apparatus and sorting method, since it is not easy to set the sorting apparatus to sort the sorting target object, the user can select a sorting apparatus that has been set in advance according to the sorting target object. I was driving with the offer. For this reason, when a mixture (waste etc.) or a sorting object is changed, there is a problem that even if the user wants to change the setting, it cannot be easily changed.

  In addition, there are various assumptions such as the case where the user's purpose of selection is to be roughly classified and the user wants to extract only a desired object with high accuracy. There was a problem that it could not be easily set or changed.

  The present invention has been made in view of such conventional problems. One of the objects of the present invention is a sorting apparatus, a sorting method, a sorting program, and a computer in which the user can easily set and change the sorting accuracy according to the purpose of sorting without special techniques and knowledge. An object is to provide a readable recording medium.

Means for Solving the Problems and Effects of the Invention

  The sorting device according to the first aspect of the present invention is a sorting device that sorts a sorting target object from a mixture composed of a plurality of types of objects, and is a type object that is the object sorted by type or A data acquisition unit that acquires data based on the mixture, a learning data generation unit that generates learning data from the data of the type object acquired by the data acquisition unit, and learning data generated by the learning data generation unit A learning unit that learns a method of classifying a mixture into types and using them as classification objects, creating a learning model in which knowledge and experience obtained by the learning are converted into data, and the selection target from the classification objects From the data of the mixture acquired by the data acquisition unit based on the learning model created in the learning unit and a selection target selection unit that selects the type of the object, each of the mixture in the mixture A first recognition rate indicating a probability that a body is a selection target selected by the selection target selection unit is calculated, and the presence / absence and position of the selection target are determined based on the first recognition rate A determination unit; and a selection unit that selects the selection object from the mixture based on a determination result of the determination unit and a threshold provided for the first recognition rate.

  According to the above configuration, during the operation of selecting the selection target from the mixture, the artificial intelligence calculates the first recognition rate indicating the probability that each object of the mixture is the selection target, and the recognition rate and the user set Since the selection target is determined by associating it with a possible threshold value, the user can control the selection accuracy while using artificial intelligence. As a result, it is possible to select according to the user's need for the selection accuracy when various purposes such as the case where the purpose of the selection only needs to be roughly classified and when it is desired to extract only a desired object with high accuracy are possible.

  In the sorting apparatus according to the second aspect of the present invention, the sorting unit can sort a sorting object having the first recognition rate equal to or higher than the threshold value. According to the said structure, it can classify | select with high precision by setting a threshold value high, and can classify | sort roughly by setting a threshold value low.

  In the sorting device according to the third aspect of the present invention, the determination unit is configured to determine each object in the mixture from the mixture data acquired by the data acquisition unit based on the learning model created by the learning unit. Calculates a second recognition rate indicating the probability of being the type object for each type object, specifies the type of each object in the mixture based on the second recognition rate, and the type is the selection The presence / absence and position of the selection object can be determined by regarding the second recognition rate in the case of matching with the type of the object as the first recognition rate. According to the above configuration, since the second recognition rate is calculated for each type of object for each object of the mixture, the type of the object can be determined as the type with the highest second recognition rate, and the selection is performed. Since an object determined to be the same type as the target object is selected by being associated with a threshold that can be set by the user, the user can select a learning model specialized for the selected target object even when the target object is changed. Therefore, it is possible to easily change the selection object.

  The sorting apparatus according to the fourth aspect of the present invention further includes a threshold setting unit that sets a desired threshold for the first recognition rate, and instructs each unit in response to an operation from a user. A data acquisition instruction unit that instructs the data acquisition unit to acquire data, and a learning data creation instruction unit that instructs the learning data creation unit to start creating the learning data. A learning start instruction unit that instructs the learning unit to create the learning model, a selection target selection instruction unit that instructs the selection of a type of the selection target to the selection target selection unit, and the threshold setting unit A threshold setting instruction unit for instructing the setting of a threshold; and causing the determination unit to determine the presence / absence and position of the selection target, and causing the selection unit to select the selection target from the mixture based on the determination result Start operation instruction When, it can be configured to have. According to the above configuration, since most of complicated setting work can be performed by artificial intelligence by simple operation, the setting for selecting a selection target by the user without specialized technology or knowledge is possible. Can be easily performed. In addition, the user can easily set / change the sorting accuracy.

  In the sorting device according to the fifth aspect of the present invention, the operation unit displays at least the data acquisition instruction unit, the learning data creation instruction unit, and the learning start instruction unit, and at least the driving start instruction unit. And a mode switching instruction unit for instructing a mode switching operation including an operation mode for displaying. According to the above-described configuration, the user can perform work while grasping which of the operation states of the selection device, that is, the learning mode and the operation mode, and the setting operation in the learning mode is Since the instruction sections related to the settings are integrated, it is easy to prevent erroneous operations.

  In the selection device according to the sixth aspect of the present invention, the operation unit includes the data acquisition instruction unit, a learning data creation instruction unit, a learning start instruction unit, a selection target selection instruction unit, a threshold setting instruction unit, and an operation start instruction unit. Can be displayed on a single screen. According to the above configuration, the learning mode and the driving mode are not distinguished as modes, and the instruction unit for setting and the instruction unit for driving are displayed on one screen. No switching operation is required.

  The sorting device according to the seventh aspect of the present invention can be configured such that the operation unit is a touch panel. According to the said structure, a user can operate easily.

  The sorting device according to the eighth aspect of the present invention can be configured such that the data acquisition unit includes a visible camera, and the data acquired by the data acquisition unit is image data. According to the said structure, since a data acquisition part is provided with a visible camera and data can be acquired as image data, a selection target can be selected based on the form, position, size, and range of the selection target. For example, when the data acquisition unit is a spectroscope-equipped camera, the data can be acquired as spectral distribution data.

  In the sorting device according to the ninth aspect of the present invention, the sorting unit can sort the sorting target from the mixture by applying compressed air to the sorting target based on the determination result.

  A sorting method according to a tenth aspect of the present invention is a sorting method for sorting a sorting target object from a mixture composed of a plurality of types of objects, and receives each operation from the data acquisition instruction unit. From the data acquisition step of acquiring data based on the classification object or the mixture that is the object classified into the object and the operation of the learning data creation instruction unit, from the data of the classification object acquired in the data acquisition step A learning data creation step for creating learning data, and a method for receiving a manipulation from the learning start instruction unit and classifying the mixture by type using the learning data created in the learning data creation step to make a type object. A learning process for learning and creating a learning model in which knowledge and experience obtained by the learning are converted into data and an operation from the selection target selection instruction unit, From the data of the mixture acquired in the data acquisition step based on the learning model created in the learning step in response to the operation from the selection target selection step of selecting the type of the selection target and the operation start instruction unit, Calculate a first recognition rate indicating the probability that each object in the mixture is a selection target selected by the selection target selection unit, and based on the first recognition rate, the presence or absence of the selection target And an operation step of determining the position and selecting the object to be selected from the mixture based on the determination result and a threshold provided for the first recognition rate.

  In the selection method according to the eleventh aspect of the present invention, in the operation step, the selection object having the first recognition rate equal to or higher than the threshold value can be selected.

  In the selection method according to the twelfth aspect of the present invention, in the operation step, each object in the mixture is obtained from the mixture data acquired in the data acquisition step based on the learning model created in the learning step. Calculates a second recognition rate indicating the probability of being the type object for each type object, specifies the type of each object in the mixture based on the second recognition rate, and the type is the selection The presence / absence and position of the selection object can be determined by regarding the second recognition rate in the case of matching with the type of the object as the first recognition rate.

  A sorting program according to the thirteenth aspect of the present invention is a sorting program for sorting a sorting target object from a mixture composed of a plurality of types of objects, and receives an operation from the data acquisition instruction unit, Creates learning data from the acquired imaging data of the type object in response to a function of acquiring data based on the type object or the mixture that is classified by type and an operation from the learning data creation instruction unit The learning function and the operation from the learning start instructing unit, learning the method of classifying the mixture into types by using the created learning data, and making the type object, and the knowledge and experience obtained by the learning A function for creating a learning model converted into data, a function for selecting the type of the selection target from the type objects in response to an operation from the selection target selection instruction unit, and an operation start instruction In response to the operation from the unit, based on the created learning model, the probability that each object in the mixture is a selection target selected by the selection target selection unit from the acquired mixture data is indicated. The first recognition rate is calculated, the presence / absence and position of the selection object is determined based on the first recognition rate, and the determination result and the threshold value provided for the first recognition rate are The computer can realize the function of selecting the object to be selected from the mixture.

  The selection program according to the fourteenth aspect of the present invention can cause a computer to realize a function of selecting a selection object having the first recognition rate equal to or higher than the threshold value.

  The selection program according to the fifteenth aspect of the present invention shows, based on the created learning model, the probability that each object in the mixture is the type object for each type object from the acquired mixture data. The second recognition rate is calculated when the second recognition rate is calculated, the type of each object in the mixture is specified based on the second recognition rate, and the type matches the type of the selection object Can be realized as the first recognition rate, and the computer can realize a function of determining the presence and position of the selection object.

  Furthermore, a computer-readable recording medium storing a program according to the sixteenth aspect of the present invention stores the program. CD-ROM, CD-R, CD-RW, flexible disk, magnetic tape, MO, DVD-ROM, DVD-RAM, DVD-R, DVD + R, DVD-RW, DVD + RW, Blu-ray (registered) Trademark), BD-R, BD-RE, HD DVD (AOD), and other magnetic disks, optical disks, magneto-optical disks, semiconductor memories, and other media that can store programs. The program includes not only a program stored in the recording medium and distributed but also a program distributed by download through a network line such as the Internet. Further, the recording medium includes a device capable of recording a program, for example, a general purpose or dedicated device in which the program is implemented in a state where it can be executed in the form of software, firmware, or the like. Furthermore, each process and function included in the program may be executed by program software that can be executed by a computer, or each part of the process and function may be performed with hardware such as a predetermined gate array (FPGA, ASIC) or program software. You may implement | achieve in the format with which the partial hardware module which implement | achieves the one part element of hardware is mixed.

It is a schematic diagram of a sorting device concerning an embodiment of the invention. It is explanatory drawing of the positional relationship of the line sensor camera which concerns on embodiment of this invention, and a conveyor. It is explanatory drawing of the image data produced with the line sensor camera which concerns on embodiment of this invention. It is explanatory drawing of the method of extracting the part which an object is imaged from image data, and producing extraction image data. It is explanatory drawing of the method of extracting the part which an object is imaged from image data, and producing extraction image data. It is explanatory drawing of the method of producing the image data of an artificial mixture. It is explanatory drawing of the method of setting the injection timing for every injection area | region and an air injection nozzle. It is a functional block diagram of the sorting device concerning an embodiment of the invention. It is a flowchart which shows the flow of the operating method of the selection apparatus in learning mode. It is explanatory drawing of the screen displayed on a controller. It is explanatory drawing of the screen displayed on a controller. It is explanatory drawing of the screen displayed on a controller. It is explanatory drawing of the screen displayed on a controller. It is explanatory drawing of the screen displayed on a controller. It is explanatory drawing of the screen displayed on a controller. It is explanatory drawing of the screen displayed on a controller. It is explanatory drawing of the screen displayed on a controller. It is a flowchart which shows the flow of the operation method of the sorting device in an operation mode. It is explanatory drawing of the screen displayed on a controller. It is explanatory drawing of the screen displayed on a controller. It is explanatory drawing of the screen displayed on a controller. It is explanatory drawing of the screen displayed on a controller.

Hereinafter, embodiments of the present invention will be described with reference to the drawings. However, the embodiment described below exemplifies a sorting device for embodying the technical idea of the present invention, and the present invention does not specify them as the following. Further, the present specification by no means specifies the members shown in the claims to the members of the embodiments. In particular, the dimensions, materials, shapes, relative arrangements, and the like of the component parts described in the embodiments are not intended to limit the scope of the present invention unless otherwise specified, and are merely explanations. It's just an example. Note that the size, positional relationship, and the like of the members shown in each drawing may be exaggerated for clarity of explanation. Furthermore, in the following description, the same name and symbol indicate the same or the same members, and detailed description thereof will be omitted as appropriate. Furthermore, each element constituting the present invention may be configured such that a plurality of elements are constituted by the same member and the plurality of elements are shared by one member, and conversely, the function of one member is constituted by a plurality of members. It can also be realized by sharing.
(Sorting device 1)

  Based on FIG. 1 which is a schematic diagram, FIG. 8 which is a functional block diagram, and FIG. 2 which is an explanatory diagram of a positional relationship between the line sensor camera 11 and the conveyor 13 for the sorting device 1 according to the embodiment of the present invention. explain.

  As shown in FIG. 1, the sorting device 1 according to the present embodiment converts a sorting object SO from a mixture MO composed of a plurality of types of objects supplied from a feeding device 2 and flowing to a conveyor 13 into compressed air. Is a device that sorts using the air injection nozzle 14 that emits the light, and mainly includes the line sensor camera 11 (corresponding to an example of “data acquisition unit” in the claims), a first control unit 12, a controller 15 ( This corresponds to an example of the “operation unit” in the claims.), The conveyor 13 and the air injection nozzle 14. The supply device 2 includes, for example, a charging hopper 21, a transfer conveyor 22, and a charging feeder 23. The charging hopper 21 is configured to receive the mixture MO. The transfer conveyor 22 supplies the mixture MO supplied from the input hopper 21 to the input feeder 23. The input feeder 23 is configured by a vibration feeder, an electromagnetic feeder, or the like, and supplies the mixture MO to the conveyor 13 while preventing the mixture MOs from overlapping with each other by vibrating.

  The sorting device 1 has two modes, a learning mode LM and an operation mode OM. The learning mode LM is a mode for performing preparation and setting for operating the sorting device 1. On the other hand, the operation mode OM is a mode for actually sorting the sorting object SO from the mixture MO.

  The mixture MO is composed of a plurality of types of objects, such as metal, paper, plastic, etc., that can identify individual objects from the image data acquired by the line sensor camera 11 and can change the course by air injection from the air injection nozzle 14. Has been. As the types of objects included in the mixture MO, for example, metal, paper, plastic, and the like are assumed, but for example, it is not limited to a large bundle such as metal, but from colors and shapes, such as copper and aluminum classified into lower layers. Anything that can be identified can be targeted. In addition, the sorting device 1 according to the present embodiment can identify up to five types of objects at a time, such as aluminum, brass, gold, silver, and copper, and is configured with such objects. From the mixture MO, it is possible to select one kind of sorting object SO, for example, only copper, or to sort several kinds at the same time, for example, aluminum, brass, and gold. Has been.

Hereinafter, each member will be described in detail. In the following description, for the sake of convenience, it is assumed that the mixture MO is composed of objects A to C (corresponding to an example of “type object” in the claims), and the object A is selected as the selection object SO. .
(Line sensor camera 11)

  As shown in FIG. 2, the sorting device 1 includes two line sensor cameras 11 arranged in the width direction of the conveyor 13. The line sensor camera 11 is a member that captures an image every time a pulse is received from the encoder 131 of the conveyor 13 and acquires an image data ID from the imaged result.

  The X direction of the line sensor camera 11 corresponds to the width direction of the conveyor 13, and the Y direction corresponds to the traveling direction of the conveyor 13. As shown in FIG. 2, a predetermined X direction imaging range 11 a can be obtained with one line sensor camera 11. Can be imaged. As shown in FIG. 3, an X-direction range 11e obtained by adding two X-direction effective ranges 11d excluding the exclusion range 11b at both ends of the conveyor 13 and the exclusion range 11c at the center of the conveyor 13 from the X-direction range 11a An image data ID is created by extracting in the Y direction within a predetermined Y direction range 11f. Among the created image data IDs, a desired overlapping range 11g from one end in the Y direction is a range overlapping with the image data ID created immediately before.

  The line sensor camera 11 in the learning mode captures an object included in the mixture MO for each object, and creates an image data ID of each object. Specifically, imaging is performed in a state where a plurality of objects A are flowed on the conveyor 13, and an image data ID of the object A is created. Similarly, the image data IDs of the objects B and C are created for the objects B and C. The created image data ID of each object is transmitted to the storage unit 121 in a state associated with the name of the imaged object. Further, imaging is performed in a state where no object is flowing on the conveyor 13, a background image BI is created, and transmitted to the storage unit 121 shown in FIG.

  In addition, the line sensor camera 11 in the operation mode OM captures an image with the mixture MO flowing on the conveyor 13, and creates an image data ID of the mixture MO. The created image data ID of the mixture MO is transmitted to the determination unit 125.

The line sensor camera 11 has been described as an example of the “data acquisition unit” in the claims. However, the “data acquisition unit” is not limited to this, and may be an area sensor camera, visible light, infrared light, or the like. , X-rays may be used. When X-rays are used, the X-ray light source can be arranged above the object conveyed by the conveyor, the X-ray camera can be arranged below the conveyor belt, and vice versa. In addition to the name of the object, the created image data ID of the object includes information that allows the user to know what kind of object the user selects when selecting the selection target SO in the selection target selection unit 124 described later. It only has to be associated. The background image BI is not necessarily captured and created by the line sensor camera 11, and the separately prepared background image BI may be stored in the storage unit 121 at the manufacturing stage of the sorting device 1.
(First control unit 12)

The first control unit 12 includes a storage unit 121, a learning data creation unit 122, a learning unit 123, a selection target selection unit 124, a threshold setting unit 126, and a determination unit 125. In the operation mode OM, the first control unit 12 determines the presence and position of the selection target object SO from the image data ID of the mixture MO acquired by the line sensor camera 11. In the learning mode LM, preparation and setting for the determination are performed. Hereinafter, each member will be described in detail.
(Storage unit 121)

The storage unit 121 is a member that stores the image data ID of the objects A to C created by the line sensor camera 11, the names of the objects associated with the image data ID, and the background image BI.
(Learning data creation unit 122)

  The learning data creation unit 122 creates and stores learning data LD from the image data ID and background image BI of the objects A to C captured and acquired by the line sensor camera 11. The learning data creation unit 122 includes three members: an image extraction unit 122a, an image composition unit 122b, and an answer creation unit 122c. The configuration of each member is as described later.

The created learning data LD is used for learning performed by the learning unit 123. One learning data LD is used for each learning, and the accuracy of selection in the operation mode OM increases as the number of repetitions of this learning increases. That is, as the learning data LD created by the learning data creation unit 122 increases, the accuracy of selection in the operation mode OM improves. The sorting device 1 according to the first embodiment of the present invention is an aspect in which the upper limit is 40,000 times and the user can freely set the number of repetitions of learning (details will be described later).
(Image extraction unit 122a)

  The image extraction unit 122a calls the image data IDs and the background images BI of the objects A to C from the storage unit 121, and extracts and extracts the portion where the object appears from the image data IDs of the objects A to C based on the background image BI. Image data SD is created. For example, when extracting the image data ID of the object A, as shown in FIG. 4, the range excluding the range overlapping with the immediately preceding image data ID is compared with the background image BI for each pixel. As a result of the comparison, a portion that does not match the background image BI is cut out as a portion where the object A appears, and the extracted image data SD of the object A is created. As described above, the comparison is basically performed in the range excluding the range overlapping with the immediately preceding image data ID. However, as shown in FIG. 5, when the object A is located outside the range, the range is overlapped. Widen the comparison. Similarly, the extracted image data SD of the objects B and C is created from the image data IDs of the objects B and C.

Note that it is not necessary to extract only a portion where an object is strictly shown, and the background image BI may remain.
(Image composition unit 122b)

  As shown in FIG. 6, the image composition unit 122b randomly selects some data from the extracted image data SD of the objects A to C created by the image extraction unit 122a, and randomly selects the background image BI. The image data ID of the artificial mixture MO is created by combining the image with the correct position, angle and size.

That is, by changing the position, angle, and size of the extracted image data SD, image data IDs of a large number of artificial mixtures MO can be created from the image data IDs of a small number of objects A to C.
(Answer creation unit 122c)

The answer creating unit 122c records information indicating which of the objects A to C is arranged in which position of the image data ID of the artificial mixture MO created by the image composition unit 122b as an image of the artificial mixture MO. Learning data LD, which is data associated with the data ID, is created.
(Learning unit 123)

  The learning unit 123 has artificial intelligence, learns a method of discriminating the objects A to C using the learning data LD created by the learning data creation unit 122, and creates a learning model GM.

Specifically, first, the probability that each object shown in the image data ID of the artificial mixture MO in the learning data LD is the object A is calculated. Similarly, the probability of being an object B and the probability of being an object C are calculated (the calculated probabilities are hereinafter referred to as recognition rates RR. The recognition rate RR is defined as “second recognition” in the claims. Corresponds to an example of "rate"). Next, each object is predicted to be the highest type of object among the recognition rates RR of the objects A to C, and it is checked whether or not the prediction has been made based on the information associated with the answer creating unit 122c. . A learning model GM, which is data obtained by repeating knowledge and experience obtained by repeating this, is created and stored.
(Selection target selection unit 124)

  The selection target selection unit 124 creates and stores a recipe RE that is data in which information on the selection target object SO selected by the user from the objects A to C is associated with the learning model GM. In the operation mode, the recipe RE selected by the user is read by the determination unit 125.

As described above, the sorting apparatus 1 is a mode in which the learning unit 123 learns the method for discriminating the objects A to C and does not learn which is the sorting object SO. Thus, for example, even when it is desired to change the selection target SO from the object A to the object B, the selection target selection unit 124 only needs to select the object B as the selection target SO. There is no need to redo learning. In addition, it is good also as an aspect which selects the selection target object SO before making the learning part 123 learn.
(Threshold setting unit 126)

The threshold value setting unit 126 sets a threshold value for the recognition rate RR of the selection object SO. Information on the set threshold value is transmitted to the second control unit 141 and is referred to when sorting the sorting object SO (details will be described later). Note that the threshold does not necessarily have to be set.
(Judgment unit 125)

  The determination unit 125 has artificial intelligence, reads the recipe RE from the selection target selection unit 124 in the operation mode OM, and the image of the mixture MO created and transmitted by the line sensor camera 11 based on the recipe RE. The presence / absence of the object A is determined from the data ID. If the object A is present, information on the position of the pixel unit is transmitted to the second control unit 141.

For the determination of the presence or absence of the object A, the recognition rate RR of the objects A to C of each object is calculated, and the object having the highest recognition rate RR of the object A is determined as the object A, as in the learning unit 123. Similarly, the object having the highest recognition rate RR of the object B is determined as the object B, and the object having the highest recognition rate RR of the object C is determined as the object C.
(Conveyor 13)

The conveyor 13 is a member that passes through the imaging range of the line sensor camera 11 and moves an object to the position of the air injection nozzle 14 to move it. The conveyor 13 moves an object at a predetermined speed. The conveyor 13 is provided with an encoder 131, and the encoder 131 transmits a pulse to the line sensor camera 11, the first controller 12, and the second controller 141 every time the conveyor 13 moves a predetermined distance. The line sensor camera 11 takes an image every time this pulse is received. That is, one pixel of the image data ID captured by the line sensor camera 11 corresponds to a predetermined distance. Moreover, the 1st control part 12 and the 2nd control part 141 pinpoint the position of an object based on this pulse.
(Air injection nozzle 14)

  The air injection nozzle 14 is a member that sorts the sorting object SO by releasing compressed air to the sorting object SO whose recognition rate RR of the sorting object SO is equal to or higher than the threshold set by the threshold setting unit 126. In the sorting device 1, a plurality of air injection nozzles 14 are arranged at minute intervals over the entire width of the conveyor 13. With the above configuration, since the selection target is determined by associating the recognition rate RR with a threshold that can be set by the user, the user can control the selection accuracy while using the artificial intelligence, and the user needs for the selection accuracy. Sorting according to this becomes possible. Specifically, if the threshold is set low, rough classification is possible, and if the threshold is set high, only a desired object can be extracted with high accuracy. The objects to be sorted are not limited to the sorting objects SO whose recognition rate RR of the sorting objects SO is equal to or higher than the threshold set by the threshold setting unit 126. For example, a selection object SO having a recognition rate RR of the selection object SO larger than a threshold set by the threshold setting unit 126 may be selected. Moreover, it is good also as an aspect which sets the threshold value of an upper limit and a minimum, and selects the selection target object SO of the recognition rate RR in the meantime, and it is good also as an aspect which selects all the selection target objects SO without setting a threshold value. . Furthermore, it is good also as an aspect which discharge | releases compressed air with respect to other than sorting object SO, and sorts sorting object SO.

  In addition, the air injection nozzle 14 is instructed by the second control unit 141 to be an injection timing that is a timing at which compressed air is injected. Specifically, as shown in FIG. 7, the second control unit first sets an injection region IR for injecting compressed air based on the position information of the object A transmitted from the determination unit 125. Next, the injection timing is set for each air injection nozzle 14 based on the injection region IR. The injection timing is provided at predetermined time intervals with respect to the traveling direction of the conveyor 13. That is, when the image data ID of the mixture MO shown in FIG. 7 is taken as an example, the air injection nozzles 14 in the rows d to h are used as a reference based on the time T0 when the upper end of the image data ID reaches the position of the air injection nozzle 14. On the other hand, the air injection nozzle 14 is instructed to inject compressed air at a timing when it passes through the injection region IR.

The object A sprayed with the compressed air by the air spray nozzle 14 is collected by the hopper 31 of the collection hopper 3 that is disposed at the lower part of the conveyor 13 and provided for each type of material to be selected. The objects B and C to which compressed air is not injected by the air injection nozzle 14 are collected by the hopper 32.
(Controller 15)

The controller 15 is a touch panel controller, and the user can easily operate the sorting device 1 by using the controller 15. The controller 15 is a mode switching button 15a (corresponding to an example of a “mode switching instruction unit” in claims), and an imaging button 15b (corresponding to an example of a “data acquisition instruction unit” in claims). Learning data creation button 15c (corresponds to an example of “learning data creation instruction unit” in claims), learning start button 15d (corresponds to an example of “learning start instruction unit” in claims) , A selection target selection button 15e (corresponding to an example of “selection target selection instruction unit” in claims), a threshold setting button 15h (corresponding to an example of “threshold setting unit” in claims), An operation start button 15f (corresponding to an example of “operation start instruction unit” in the claims) and an operation end button 15g are provided.
(Operation method of the sorting apparatus 1)

Hereinafter, a method of operating the sorting device 1 using the controller 15 will be described.
(Operation method in learning mode LM)

  The operation method of the sorting apparatus 1 in the learning mode LM will be described based on the functional block diagram of FIG. 8, the flowchart of FIG. 9, and the explanatory diagrams of the screens displayed on the controller 15 of FIGS.

  First, in step ST101, the selection device 1 is switched to the learning mode LM using the mode switching button 15a. When the sorting device 1 is activated, the screen shown in FIG. 10 is displayed on the controller 15, so that the sorting device 1 can be switched to the learning mode LM by pressing the learning mode button 151a, and the controller 15 shows the screen shown in FIG. A screen is displayed.

  Next, in step ST102, the line sensor camera 11 is caused to create image data IDs and background images BI of the objects A to C. When the user flows a plurality of objects A onto the conveyor and presses the imaging button 15b on the screen shown in FIG. 11, the line sensor camera 11 starts imaging and creates an image data ID of the object A. When the acquisition of the image data ID of the object A is completed, the screen shown in FIG. 12 is displayed on the controller 15, and the user inputs the name of the object A to the name input unit 151 b and stores it in the storage unit 121. When the storage of the object A is completed, the screen of FIG. 11 is displayed again on the controller 15, so that the user captures the objects B and C and the background image BI in the same procedure.

  Next, in step ST103, the learning data creation unit is caused to create learning data LD. When the user presses the learning data creation button 15 c on the screen shown in FIG. 11, the screen shown in FIG. 13 is displayed on the controller 15. When the user presses the object selection button 151c, an object (in the case of this description, “ Object A "," Object B "," Object C "). When the selection is completed, the screen shown in FIG. 13 is displayed again on the controller 15, and the number of learning data LD to be created is input to the data number input unit 152c. When the input is completed, as shown in FIG. 15, the controller 15 displays a standby screen indicating the expected time until the creation of the learning data LD is completed for the user. When the creation of the learning data LD is completed, the controller 15 displays a screen shown in FIG.

Finally, in step ST104, the learning unit 123 is trained using the learning data LD to create a learning model GM. When the user presses the learning start button 15 d on the screen shown in FIG. 11, the screen shown in FIG. 16 is displayed on the controller 15. The user learns from the list of learning data LD stored in the learning data creation unit 122 displayed as shown in FIG. 16 (the name of the object used to create the learning data LD is displayed). Learning data LD (in the case of this description, “object A, object B, object C”) is selected. When the selection is completed, as shown in FIG. 17, the controller 15 displays a standby screen indicating the expected time until the creation of the learning model GM is completed. When the creation of the learning model GM is completed, the controller 15 displays a screen shown in FIG.
(Operation method in operation mode OM)

  The operation method of the sorting apparatus 1 in the operation mode OM will be described based on the functional block diagram of FIG. 8, the flowchart of FIG. 18, and the explanatory diagrams of the screens displayed on the controller 15 of FIGS. 19 to 21.

  First, in step ST201, the sorting device 1 is switched to the operation mode OM using the mode switching button 15a. Since the screen shown in FIG. 10 is displayed on the controller 15 when the sorting device 1 is activated, the sorting device 1 is switched to the operation mode OM by pressing the operation mode button 152a, and the controller 15 shows the screen shown in FIG. A screen is displayed.

  Next, in step ST202, the object A is selected as the sorting object SO, and the sorting target selecting unit 124 is caused to create a recipe RE. When the user presses the selection target selection button 15e on the screen shown in FIG. 19, the controller 15 displays the screen shown in FIG. The user uses a learning model used for discrimination from a list of learning models GM stored in the learning unit 123 displayed as shown in FIG. 20 (the names of objects used to create the learning model GM are displayed). GM ("Object A, Object B, Object C" in this embodiment) is selected. When the selection is completed, the controller 15 displays a screen shown in FIG. The user selects the sorting object SO (“object A” in this description) from the list of objects used to create the selected learning model GM displayed as shown in FIG. When the selection is completed, the selection target selection unit 124 creates a recipe RE, and the controller 15 displays a screen shown in FIG.

  Next, in step ST203, a threshold is set for the recognition rate RR of the selection object SO. When the user presses the threshold setting button 15h on the screen shown in FIG. 19, the screen shown in FIG. 22 is displayed on the controller 15, and a desired threshold is input to the threshold input unit 151h. When the input is completed, the threshold value setting unit 126 transmits threshold value information to the second control unit 141, and the controller 15 displays the screen shown in FIG. When a desired threshold value is not input to the threshold value input unit 151h, it is determined that no threshold value has been set, and all the selection objects SO are selected.

  Next, the object A is selected in step ST204. When the user flows the mixture MO on the conveyor and presses the operation start button 15f on the screen shown in FIG. 19, the line sensor camera 11 starts imaging, and the determination unit 125 determines whether the object A is present and whether the object A is in pixel units. The position is determined, and the air injection nozzle 14 selects the object A based on this determination.

  Finally, in step ST205, the operation end button 15g is pressed to finish the selection.

  Note that the aspect of the controller 15 and the screen display are not limited to those described above, and may be changed as appropriate so that the user can easily operate the sorting device 1. For example, the controller 15 using a push button may be used, and in this case, the mode switching button 15a is unnecessary. Moreover, it is good also as an aspect which displays all the buttons on one screen, without providing the mode switching button 15a. Further, the controller 15 may be displayed to instruct the user to perform the next operation.

  In the above-described embodiment, each button has a separate function. However, each function may be linked, or a predetermined button may have various functions. For example, the learning data LD may be created by pressing the learning data creation button 15c, and the learning model GM may be created based on the learning data. Further, for example, the operation start button 15f may also have a function of instructing the end of operation, and the operation may be started by pressing the first operation start button 15f, and the operation may be ended by pressing the second time. In the above-described embodiment, the object A is described as the selection target. However, a plurality of objects may be selected as the selection target, and a plurality of air injection nozzles and hoppers may be provided in accordance with the selection.

  As described above, the sorting apparatus 1 to which the present invention is applied can determine the presence and position and the position of the sorting object SO from the imaging data of the mixture MO using artificial intelligence. In addition, the recognition rate indicating the probability that each object of the mixture is a selection target is calculated by the artificial intelligence, and the selection target is determined by associating the recognition rate with a threshold that can be set by the user. As a result, the user can control the sorting accuracy.

  Moreover, it can operate easily including the process of setting a threshold value using the various buttons displayed on the controller 15. Therefore, according to the present invention, since most of complicated setting work can be performed by artificial intelligence by a simple operation, the user sorts the sorting object SO without any specialized technique or knowledge. Can be set easily.

  The sorting apparatus, sorting method, sorting program, and computer-readable recording medium according to the present invention can be applied to applications for sorting objects into two or more types.

DESCRIPTION OF SYMBOLS 1 ... Sorting device 11 ... Line sensor camera 11a ... X direction imaging range; 11b, 11c ... Exclusion range; 11d ... X direction effective range; 11e ... X direction range; 11f ... Y direction range; Control unit 121 ... storage unit 122 ... learning data creation unit; 122a ... image extraction unit; 122b ... image composition unit; 122c ... answer creation unit 123 ... learning unit 124 ... type selection unit 126 ... threshold setting unit 125 ... determination unit 13 ... Conveyor 131 ... Encoder 14 ... Air injection nozzle 141 ... Second control unit 15 ... Controller 15a ... Mode switching button; 151a ... Learning mode button; 152a ... Operation mode button 15b ... Imaging button; 151b ... Name input unit 15c ... Learning data creation 151c ... Object selection button; 152c ... Number of data input section 15d ... Learning start button 15 Selection target selection button 15h ... Threshold setting button; 151h ... Threshold input unit 15f ... Operation start button 15g ... Operation end button 2 ... Supply device 21 ... Input hopper; 22 ... Transfer conveyor; 23 ... Input feeder 3 ... Collection hopper 31, 32 ... hopper MO ... mixture SO ... selection object ID ... image data LD ... learning data SD ... extracted image data BI ... background image GM ... learning model RE ... recipe RR ... recognition rate IR ... injection region LM ... learning mode OM ... driving mode

Claims (16)

A sorting device that sorts a sorting target from a mixture of a plurality of types of objects,
A data acquisition unit that acquires data based on the type object or the mixture that is the object sorted by type; and
A learning data creation unit that creates learning data from the data of the type object acquired by the data acquisition unit;
Learning to classify a mixture by type using the learning data created by the learning data creation unit, learn how to make a type object, and create a learning model that converts the knowledge and experience obtained by the learning into data And
A selection target selection unit for selecting the type of the selection target from the type objects;
Based on the learning model created by the learning unit, from the mixture data acquired by the data acquisition unit, the probability that each object in the mixture is a selection target selected by the selection target selection unit A determination unit that calculates a first recognition rate to indicate, and based on the first recognition rate, determines the presence and position of the selection object;
A sorting unit that sorts the sorting object from the mixture based on a judgment result of the judging unit and a threshold value set for the first recognition rate;
A sorting apparatus comprising: The sorting device according to claim 1,
The sorting apparatus, wherein the sorting unit sorts a sorting object having the first recognition rate equal to or higher than the threshold value. The sorting apparatus according to claim 1 or 2,
The determination unit determines, based on the learning model created by the learning unit, the probability that each object in the mixture is the type object for each type object from the mixture data acquired by the data acquisition unit. A second recognition rate when the second recognition rate is calculated, the type of each object in the mixture is specified based on the second recognition rate, and the type matches the type of the selection object A sorting apparatus characterized in that a rate is regarded as the first recognition rate, and the presence and position of the sorting object are determined. The sorting apparatus according to any one of claims 1 to 3, further comprising:
A threshold setting unit for setting a desired threshold for the first recognition rate;
An operation unit that gives an instruction in response to an operation from the user with respect to each unit,
With
The operation unit is
A data acquisition instruction unit that instructs the data acquisition unit to acquire data;
A learning data creation instruction unit that instructs the learning data creation unit to start creating the learning data;
A learning start instruction unit that instructs the learning unit to create the learning model;
A selection target selection instruction unit that instructs the selection target selection unit to select the type of the selection target;
A threshold setting instruction unit for instructing the threshold setting unit to set the threshold;
An operation start instruction unit that causes the determination unit to determine the presence and position of the selection target, and causes the selection unit to select the selection target from the mixture based on the determination result;
A sorting apparatus comprising: The sorting device according to claim 4,
The operation unit instructs a mode switching operation including at least a learning mode for displaying the data acquisition instruction unit, a learning data creation instruction unit, and a learning start instruction unit, and at least an operation mode for displaying the operation start instruction unit. A sorting apparatus comprising a mode switching instruction unit. The sorting device according to claim 4,
The selection device, wherein the operation unit displays the data acquisition instruction unit, the learning data creation instruction unit, the learning start instruction unit, the selection target selection instruction unit, the threshold setting instruction unit, and the operation start instruction unit on one screen. . The sorting device according to any one of claims 4 to 6,
The selection device, wherein the operation unit is a touch panel. The sorting device according to any one of claims 1 to 7,
The data acquisition unit includes a visible camera;
The sorting apparatus characterized in that the data acquired by the data acquisition unit is image data. It is the sorting device according to any one of claims 1 to 8,
The sorting device, wherein the sorting unit sorts the sorting object from the mixture by applying compressed air to the sorting object based on the determination result. A sorting method for sorting a sorting object from a mixture composed of a plurality of types of objects,
In response to an operation from the data acquisition instruction unit, a data acquisition step of acquiring data based on the type object or the mixture that is the object sorted for each type;
In response to an operation from the learning data creation instruction unit, a learning data creation step of creating learning data from the data of the type object acquired in the data acquisition step;
In response to an operation from the learning start instruction unit, the learning data created in the learning data creation step is used to classify the mixture into types and learn how to make a type object, and the knowledge obtained by the learning and A learning process to create a learning model that converts experience into data,
In response to an operation from the selection target selection instruction unit, a selection target selection step of selecting the type of the selection target from the type objects;
In response to an operation from the operation start instruction unit, based on the learning model created in the learning step, each object in the mixture is determined by the selection target selection unit from the mixture data acquired in the data acquisition step. Calculating a first recognition rate indicating the probability of being a selected selection object, determining the presence and position of the selection object based on the first recognition rate, and determining the determination result and the first An operation step of selecting the selection object from the mixture based on a threshold value set for a recognition rate;
Sorting method. The screening method according to claim 10,
In the operation step, a sorting method in which the sorting object having the first recognition rate equal to or higher than the threshold is sorted. The screening method according to claim 10 or claim 11,
In the driving step, based on the learning model created in the learning step, the probability that each object in the mixture is the type object for each type object from the mixture data acquired in the data acquisition step. A second recognition rate when the second recognition rate is calculated, the type of each object in the mixture is specified based on the second recognition rate, and the type matches the type of the selection object A sorting method characterized in that a rate is regarded as the first recognition rate, and the presence / absence and position of the sorting object are determined. A selection program for selecting a selection target from a mixture composed of a plurality of types of objects,
A function of receiving data based on the type object or the mixture, which is the object sorted by type, in response to an operation from the data acquisition instruction unit;
In response to an operation from the learning data creation instruction unit, a function of creating learning data from the acquired imaging data of the type object;
Learning that receives the operation from the learning start instruction unit, classifies the mixture by type using the created learning data, learns the method of making the type object, and converts the knowledge and experience obtained by the learning into data The ability to create models,
In response to an operation from the selection target selection instruction unit, a function of selecting the type of the selection target from the type objects;
In response to an operation from the operation start instruction unit, based on the created learning model, each object in the mixture is a selection target selected by the selection target selection unit from the acquired mixture data. A first recognition rate indicating a probability is calculated, and based on the first recognition rate, the presence / absence and position of the selection object are determined, and a threshold provided for the determination result and the first recognition rate is determined. A function to sort the sorting object from the mixture based on
A selection program that enables computers to realize this. The screening program according to claim 13,
A selection program for selecting a selection object having the first recognition rate equal to or higher than the threshold. The screening program according to claim 13 or 14,
Based on the created learning model, the second recognition rate indicating the probability that each object in the mixture is the type object for each type object is calculated from the acquired data of the mixture, and the second Based on the recognition rate, the type of each object in the mixture is specified, and the second recognition rate when the type matches the type of the selection object is regarded as the first recognition rate, and the sorting is performed. A screening program characterized by determining the presence and position of an object.   A computer-readable recording medium or a device that stores the program according to any one of claims 13 to 15.

Images