KR102572455B1 - Apparatus and method for predicting the weight of mixed scrap metal waste based on image analysis - Google Patents

Abstract

The present invention relates to a device and a method for predicting a mixed scrap metal waste weight based on image analysis. According to an embodiment of the present invention, an electronic device comprises a memory and a processor connected to the memory. The processor can receive a first image from a camera installed to photograph a transport means loading a mixed scrap metal waste in a perspective view along a movement line of the transport means loading the mixed scrap metal waste, and generating the first image of the mixed scrap metal and the transport means, generate mixing information including the metal type and metal ratio of a metal included in the mixed scrap metal waste by analyzing the first image, generate weight information on the mixed scrap metal waste by analyzing the first image, and derive transaction costs of the mixed scrap metal waste on the basis of the mixing information and the weight information.

Description

Apparatus and method for predicting mixed scrap waste weight based on image analysis {APPARATUS AND METHOD FOR PREDICTING THE WEIGHT OF MIXED SCRAP METAL WASTE BASED ON IMAGE ANALYSIS}

The present invention relates to an apparatus and method for predicting the weight of mixed scrap waste based on image analysis.

delete

Unless otherwise indicated herein, material described in this section is not prior art to the claims in this application, and inclusion in this section is not an admission that it is prior art.

Metals such as scrap metal can be recycled, and are subject to purchase at junk shops or recycling centers. In the case of a single type, the purchase price can be determined only by the unit price and weight of the metal, but in the case of mixed scrap metal in which various types are mixed, it may be difficult to simply determine the purchase price.

When purchasing mixed scrap metal waste, the purchase amount must be settled immediately, but it is difficult to wait until a separate sorting process for recycling is completed.

Accordingly, the present invention intends to propose a technique capable of deriving transaction costs by predicting the weight of mixed scrap metal and predicting the unit price of mixed scrap metal based on image analysis.

delete

Korean Patent Registration No. 10-1017248 (2011.02.17.)

An embodiment of the present invention is to provide an apparatus and method for predicting the weight of mixed scrap waste based on image analysis.

The technical problems to be achieved in the present invention are not limited to the technical problems mentioned above, and other technical problems not mentioned will be clearly understood by those skilled in the art from the description below. You will be able to.

delete

In order to achieve the above object, an electronic device according to an embodiment of the present invention includes a memory and a processor connected to the memory, and the processor is connected to a moving line of a means of transportation loaded with mixed scrap metal waste. Accordingly, the first image is received from a camera that is installed to photograph the transport vehicle carrying the mixed scrap metal waste in a perspective view direction, and the first image is received from a camera that generates a first image of the mixed scrap metal waste and the transport vehicle, and the first image is analyzed. , Mixed information including metal types and metal ratios of metals included in the mixed scrap waste is generated, and weight information of the mixed scrap waste is generated by analyzing the first image, and the mixed information and the weight information are generated. Based on , the transaction cost of the mixed scrap waste can be derived.

At this time, the processor, through an artificial intelligence module, extracts only the mixed scrap metal waste from the first image to generate a second image, and corrects the mixed scrap waste covered by the transportation means in the second image. to generate a third image, and to generate the weight information based on the third image.

At this time, the processor identifies an accommodation space accommodating the mixed scrap waste in the second image through an artificial intelligence module, and among the identified accommodation spaces, a rear plate positioned on the rear side of the vehicle; The side plate located on the side of the vehicle is divided, the lower line of the rear plate is set as the first lower line, the lower line of the side plate is set as the second lower line, and the most A point located on the left is set as a first point, a point located on the rightmost side of the second image is set as a second point, and a first vertical line vertically descended from the first point is set to the first lower line. and connects a second vertical line vertically lowered from the second point to the second lower line, and connects the second image and the first vertical line, the second vertical line, the first lower line, and the second lower line. It is possible to generate the third image, including.

At this time, the processor generates the volume information of the mixed scrap metal waste based on the third image, and determines the volume information of the mixed scrap waste based on a DB including specific gravity information for each metal type, the volume information, and the mixed information. Weight information can be generated.

At this time, the processor compares the lengths of the first vertical line and the second vertical line based on the third image, and if the difference between the first vertical line and the second vertical line is within a preset error range, the preset The volume information is generated in a first type for generating the volume information based on a first volume value of a first three-dimensional object of a first shape and a second volume value of a second three-dimensional object of a predetermined second shape, and If the difference between the vertical line and the second vertical line is out of the error range, the second volume value, the third volume value of the third solid object of the first shape, the fourth volume value of the fourth solid object of the second shape, The volume information may be generated in a second type that generates the volume information based on the fifth volume value of the fifth three-dimensional object of the second shape.

At this time, in the first type, the average of the first vertical line and the second vertical line is set as a first height value, the length of the first lower line is set as a horizontal value, and the length of the second lower line is set as a vertical value. set to a value, the first shape is a rectangular parallelepiped, the first volume value of the first three-dimensional object is derived based on the first height value, the horizontal value, and the vertical value, and the uppermost A point located at is set as a top point, a tower vertical line descending vertically from the top point is created, a tower reference line connecting the first point and the second point is created, and the tower vertical line and the tower reference line are created. A length between the center point and the top point where the center point and the top point meet is set as a second height value, the second shape is a quadrangular pyramid, and based on the second height value, the horizontal value, and the vertical value, the second height value A type of deriving a second volume value and generating the volume information through the following equation,

VI (Volume Information) may mean the volume information, VV (Value of Volume)_1 may mean the first volume value, and VV_2 may mean the second volume value.

In this case, in the second type, a shorter line among the first vertical line and the second vertical line is set as the third vertical line, the length of the third vertical line is set as a third height value, and the third height is set. value, the third volume value of the third solid object is derived based on the horizontal value and the vertical value, the length difference value between the first vertical line and the second vertical line is set as a fourth height value, and the 4 The fourth volume value of the fourth solid object is derived based on the height value, the horizontal value, and the vertical value, and the fifth height of the fifth solid object is derived based on the second height value and the fourth height value. A type of deriving a value, deriving the fifth volume value of the fifth solid object based on the fifth height value, the horizontal value, and the vertical value, and generating the volume information through the following equation,

VI denotes the volume information, VV_2 denotes the second volume value, VV_3 denotes the third volume value, VV_4 denotes the fourth volume value, and VV_5 denotes the fifth volume value. can mean

At this time, the fifth height value is derived by the following equation,

H(Height)_5 may mean the fifth height value, H_2 may mean the second height value, and H_4 may mean the fourth height value.

At this time, the processor derives the average specific gravity of the mixed scrap waste based on the DB including the specific gravity information for each metal type and the metal type and the metal ratio included in the mixed information, and the volume information and the average The weight information may be derived based on the specific gravity.

At this time, the processor derives the mixed unit price information of the mixed scrap waste based on the metal ratio and the unit price per unit weight preset for each metal type, and derives the transaction cost based on the mixed unit price information and the weight information. can do.

delete

As such, according to an embodiment of the present invention, an apparatus and method for predicting the weight of mixed scrap waste based on image analysis can be provided.

The effects obtainable in the present invention are not limited to the effects mentioned above, and other effects not mentioned can be clearly understood by those skilled in the art from the description below. will be.

delete

Also other aspects as described above, features and benefits of certain preferred embodiments of the present invention will become more apparent from the following description taken in conjunction with the accompanying drawings.
1 is a conceptual diagram of an apparatus for predicting the weight of mixed scrap waste based on image analysis according to an embodiment of the present invention.
2 is a block diagram of an electronic device according to an embodiment of the present invention.
3 is a diagram illustrating generation of a first image, a second image, and a third image according to an embodiment of the present invention.
4 is a diagram illustrating generation of a third image according to an embodiment of the present invention.
5 is a diagram illustrating generation of volume information according to a first type and a second type according to an embodiment of the present invention.
6 is a diagram illustrating derivation of a fifth height value according to an embodiment of the present invention.
7 is a diagram illustrating extraction of a small object and a large object according to an embodiment of the present invention.
8 is a diagram showing an operation outline of a similarity module according to an embodiment of the present invention.
9 is a conceptual diagram for deriving classification weights according to an embodiment of the present invention.
10 is a flowchart of a method for predicting the weight of mixed scrap waste based on image analysis according to an embodiment of the present invention.
It should be noted that throughout the drawings, like reference numbers are used to show the same or similar elements, features and structures.

Hereinafter, embodiments of the present invention will be described in detail with reference to the accompanying drawings.

In describing the embodiments, descriptions of technical contents that are well known in the technical field to which the present invention pertains and are not directly related to the present invention will be omitted. This is to more clearly convey the gist of the present invention without obscuring it by omitting unnecessary description.

For the same reason, in the accompanying drawings, some components are exaggerated, omitted, or schematically illustrated. Also, the size of each component does not entirely reflect the actual size. In each figure, the same reference number is assigned to the same or corresponding component.

Advantages and features of the present invention, and methods for achieving them, will become clear with reference to the embodiments described below in detail in conjunction with the accompanying drawings. However, the present invention is not limited to the embodiments disclosed below, but may be implemented in various different forms, and only the present embodiments make the disclosure of the present invention complete, and common knowledge in the art to which the present invention belongs It is provided to fully inform the holder of the scope of the invention, and the present invention is only defined by the scope of the claims. Like reference numbers designate like elements throughout the specification.

At this time, it will be understood that each block of the process flow chart diagrams and combinations of the flow chart diagrams can be performed by computer program instructions. These computer program instructions may be embodied in a processor of a general purpose computer, special purpose computer, or other programmable data processing equipment, so that the instructions executed by the processor of the computer or other programmable data processing equipment are described in the flowchart block(s). It creates means to perform functions. These computer program instructions may also be stored in a computer usable or computer readable memory that can be directed to a computer or other programmable data processing equipment to implement functionality in a particular way, such that the computer usable or computer readable memory The instructions stored in are also capable of producing an article of manufacture containing instruction means that perform the functions described in the flowchart block(s). The computer program instructions can also be loaded on a computer or other programmable data processing equipment, so that a series of operational steps are performed on the computer or other programmable data processing equipment to create a computer-executed process to generate computer or other programmable data processing equipment. Instructions for performing processing equipment may also provide steps for performing the functions described in the flowchart block(s).

Additionally, each block may represent a module, segment, or portion of code that includes one or more executable instructions for executing specified logical function(s). It should also be noted that in some alternative implementations it is possible for the functions mentioned in the blocks to occur out of order. For example, two blocks shown in succession may in fact be executed substantially concurrently, or the blocks may sometimes be executed in reverse order depending on their function.

At this time, the term '~unit' used in this embodiment means software or a hardware component such as a field-programmable gate array (FPGA) or application specific integrated circuit (ASIC), and what role does '~unit' have? perform them However, '~ part' is not limited to software or hardware. '~bu' may be configured to be in an addressable storage medium and may be configured to reproduce one or more processors. Therefore, as an example, '~unit' refers to components such as software components, object-oriented software components, class components, and task components, processes, functions, properties, and procedures. , subroutines, segments of program code, drivers, firmware, microcode, circuitry, data, databases, data structures, tables, arrays, and variables. Functions provided within components and '~units' may be combined into smaller numbers of components and '~units' or further separated into additional components and '~units'. In addition, components and '~units' may be implemented to play one or more CPUs in a device or a secure multimedia card.

In describing the embodiments of the present invention in detail, an example of a specific system will be the main target, but the main subject matter to be claimed in this specification extends the scope disclosed herein to other communication systems and services having a similar technical background. It can be applied within a range that does not deviate greatly, and this will be possible with the judgment of those skilled in the art.

1 is a conceptual diagram of an apparatus for predicting the weight of mixed scrap waste based on image analysis according to an embodiment of the present invention.

Referring to FIG. 1 , an image analysis-based mixed scrap metal weight prediction device according to an embodiment of the present invention generates an image of mixed scrap waste through a camera, analyzes the image, and analyzes the image included in the mixed scrap waste It is possible to derive the types of metals and the proportions of each type of metal, and derive the weight information of the mixed scrap waste to derive the trading cost, ie, the transaction cost, of the mixed scrap waste.

Meanwhile, the device for predicting the weight of mixed scrap waste based on image analysis may be referred to as 'electronic device 100' in the present invention.

delete

2 is a block diagram of an electronic device 100 according to an embodiment of the present invention.

An electronic device 100 according to an embodiment includes a processor 110 and a memory 120 . The processor 110 may perform at least one method described above. The memory 120 may store information related to the above method or store a program in which the above method is implemented. Memory 120 may be volatile memory or non-volatile memory. The memory 120 may be referred to as a 'database', a 'storage unit', or the like.

The processor 110 may execute a program and control the electronic device 100 . Program codes executed by the processor 110 may be stored in the memory 120 . The device 100 may be connected to an external device (eg, a personal computer or network) through an input/output device (not shown) and exchange data.

At this time, the processor 110 is installed so that the transport means loaded with mixed scrap metal waste is photographed in a perspective direction along the movement of the transport means loaded with mixed scrap metal waste, and the first image of the mixed scrap waste and the transport means is taken. The first image may be received from a camera that generates a .

This is to create an image in which the mixed scrap waste can be clearly seen as much as possible.

In addition, the processor 110 may analyze the first image to generate mixed information including metal types and metal ratios of metals included in the mixed scrap metal waste.

The metal type may be a metal type previously set by a user, and may be, for example, iron, copper, zinc, or aluminum.

In addition, as described later, the metal ratio may mean a relative ratio of metal types included in the mixed scrap waste. For example, when iron, copper, and zinc are included in the mixed scrap iron waste, the metal types may be iron, copper, and zinc, and the metal ratio is 5:2:1 in order of iron, copper, and zinc. can be set. A method of deriving the metal type and the metal ratio will be described later in detail.

Also, the processor 110 may generate weight information of the mixed scrap waste by analyzing the first image. Regarding the generation of weight information, it will be described later with reference to drawings.

Also, the processor 110 may derive a transaction cost of the mixed scrap waste based on the mixed information and the weight information.

To this end, the processor 110 may derive mixed unit price information for each predetermined weight of the mixed scrap waste based on the mixed information.

The mixed unit price information may mean a unit price per unit weight of the mixed scrap metal waste. For example, when the mixed scrap metal waste is 1 kg, it may be set to 1000 won or the like. A process of generating the mixed unit price information will be described later in detail.

Also, the processor 110 may calculate a transaction cost of the mixed scrap waste based on the weight information and the mixed unit price information. For example, when the weight information is 100 kg and the combined unit price information is 1000 won per 1 kg, the transaction cost may be calculated as 100 * 1000 = 100000 won.

delete

3 is a diagram illustrating generation of a first image, a second image, and a third image according to an embodiment of the present invention.

The first image generated by the camera may include the transportation means and the mixed scrap waste together. According to the object of the present invention, since an image including only the mixed scrap metal waste should be analyzed, it is necessary to extract only the mixed scrap metal waste from the first image.

Accordingly, the processor 110 may generate a second image by extracting only the mixed scrap waste from the first image through an artificial intelligence module.

More specifically, the artificial intelligence module may create the second image by generating an outline of the mixed scrap waste from the first image and extracting only the image of the mixed scrap waste included in the outline.

At this time, the artificial intelligence module may use a deep learning technique, which is a field of machine learning.

In addition, the artificial intelligence module may calculate weights of a plurality of inputs in the function through deep learning through learning. In addition, various models such as RNN (Recurrent Neural Network), DNN (Deep Neural Network), and DRNN (Dynamic Recurrent Neural Network) can be used as artificial intelligence network models used for such learning.

Here, RNN is a deep learning technique that simultaneously considers current data and past data, and recurrent neural network (RNN) represents a neural network in which connections between units constituting an artificial neural network constitute a directed cycle. Furthermore, various methods can be used for the structure capable of configuring the recurrent neural network (RNN), for example, a Fully Recurrent Network, a Hopfield Network, an Elman Network, ESN (Echo state network), long short term memory network (LSTM), bi-directional RNN, continuous-time RNN (CTRNN), hierarchical RNN, and quadratic RNN are representative examples. In addition, as a method for learning a recurrent neural network (RNN), methods such as gradient descent, Hessian Free Optimization, and Global Optimization Method may be used.

At this time, the mixed scrap metal waste may be in a state in which various metals are separated or not arranged. Therefore, it is necessary to identify all objects in the second image and analyze the relationship with each object to determine whether it is one metal or not, and through this, the type and ratio of the metal can be derived. In this regard, it will be described later in detail with reference to FIG. 7 .

In addition, the processor 110 may generate a third image by correcting the mixed scrap waste covered by the transportation means in the second image in order to correct the mixed scrap waste covered by the transportation means. .

Also, the processor 110 may generate the weight information based on the third image.

delete

4 is a diagram illustrating generation of a third image according to an embodiment of the present invention.

Referring to FIG. 4, an artificial intelligence module identifies an accommodation space accommodating the mixed scrap waste in the second image, and among the identified accommodation spaces, a rear plate located on the rear side of the vehicle; Separate the side plate located on the side of the vehicle, set the lower line of the rear plate as the first lower line, set the lower line of the side plate as the second lower line, and set the leftmost line in the second image A point located at is set as a first point, a point located at the far right in the second image is set as a second point, and a first vertical line vertically descended from the first point is connected to the first lower line. and connects a second vertical line vertically descended from the second point to the second lower line, and connects the second image and the first vertical line, the second vertical line, the first lower line, and the second lower line. Including, the third image may be generated.

Through this, it is possible to derive the volume of the mixed scrap waste covered by the accommodation space of the transportation means.

In addition, the processor 110 may generate the volume information of the mixed scrap metal waste based on the third image, and based on a DB including specific gravity information for each metal type, the volume information, and the mixed information, It is possible to generate weight information of mixed scrap metal waste.

In this case, the specific gravity is information about weight per unit volume, and the DB including the specific gravity information for each metal type may include each specific gravity according to the metal type that may be included in the mixed scrap iron waste.

delete

5 is a diagram illustrating generation of volume information according to a first type and a second type according to an embodiment of the present invention.

Referring to FIG. 5 , the processor 110 derives the type of three-dimensional shape to be measured according to the shape of the third image, and derives the volume of the mixed scrap waste based on the approximate volume of each three-dimensional shape. . At this time, since measuring the accurate volume may require a lot of load and time on the system, it is to quickly derive the volume information from the three-dimensional shape.

At this time, the processor 110 compares the lengths of the first vertical line and the second vertical line based on the third image, and if the difference between the first vertical line and the second vertical line is within a predetermined error range, generating the volume information in a first type based on a first volume value of a first three-dimensional object of a preset first shape and a second volume value of a second three-dimensional object of a second preset shape; If the difference between the first vertical line and the second vertical line is out of the error range, the second volume value, the third volume value of the third solid object of the first shape, and the fourth volume value of the fourth solid object of the second shape The volume information may be generated in a second type that generates the volume information based on the volume value and the fifth volume value of the fifth three-dimensional object of the second shape.

In this case, the first shape may be a rectangular parallelepiped, and the second shape may be a quadrangular pyramid.

At this time, the processor 110, when the lengths of the first vertical line and the second vertical line are similar, may be simplified into a rectangular parallelepiped having the first vertical line and the second vertical line. And, by simplifying the rectangular parallelepiped to a quadrangular pyramid having the highest protruding point as a vertex, the volume information of the mixed scrap waste can be derived as the sum of the volumes of the rectangular parallelepiped and the quadrangular pyramid.

In addition, if the lengths of the first vertical line and the second vertical line are not similar, the processor 110 first simplifies it into a rectangular parallelepiped having a short line, and then simplifies it into a first quadrangular pyramid having a point having a long line as a vertex. can In addition, volume information of the mixed scrap waste may be derived by further simplifying a second quadrangular pyramid having the highest protruding point as a vertex, and removing an overlapping region between the first quadrangular pyramid and the second quadrangular pyramid.

Looking in more detail, in the first type, the average of the first vertical line and the second vertical line is set as a first height value, the length of the first lower line is set as a horizontal value, and the length of the second lower line is set. is set as a vertical value, the first shape is a rectangular parallelepiped, and the first volume value of the first three-dimensional object may be derived based on the first height value, the horizontal value, and the vertical value.

In this case, the first volume value may be derived as a horizontal value * a vertical value * a first height value according to the volume formula of a rectangular parallelepiped.

The length may be derived by a corresponding equation between the length in the third image and the actual length based on the position and angle of the camera and the position of the vehicle.

In addition, in the first type, a point located at the top of the third image is set as a top point, a top vertical line is created that descends vertically from the top point, and the first point and the second point are connected. A tower reference line is created, and a length between a center point where the tower vertical line and the tower reference line meet and the top point is set as a second height value, the second shape is a quadrangular pyramid, the second height value, the width The second volume value of the second three-dimensional object may be derived based on the value and the vertical value.

In addition, the first type may be a type for generating the volume information through Equation 1 below.

[Equation 1]

In this case, VI (Volume Information) may mean the volume information, VV (Value of Volume)_1 may mean the first volume value, and VV_2 may mean the second volume value.

In addition, in the second type, a shorter line among the first vertical line and the second vertical line is set as a third vertical line, the length of the third vertical line is set as a third height value, and the third height value , The third volume value of the third solid object may be derived based on the horizontal value and the vertical value.

In addition, in the second type, a length difference between the first vertical line and the second vertical line is set as a fourth height value, and the fourth three-dimensional object is formed based on the fourth height value, the horizontal value, and the vertical value. The fourth volume value may be derived.

In addition, the second type derives a fifth height value of the fifth three-dimensional object based on the second height value and the fourth height value, and based on the fifth height value, the horizontal value, and the vertical value The fifth volume value of the fifth solid object may be derived.

In this case, the second three-dimensional object and the fourth three-dimensional object have the shape of a quadrangular pyramid, and the overlapping area can be simplified to a quadrangular pyramid shape. Accordingly, the fifth height value may be extracted based on the second height value and the fourth height value, and the entire volume information may be derived by removing the quadrangular pyramid having the fifth height value as an overlapping quadrangular pyramid.

At this time, the second type may be a type that generates the volume information through Equation 2 below.

[Equation 2]

At this time, VI means the volume information, VV_2 means the second volume value, VV_3 means the third volume value, VV_4 means the fourth volume value, and VV_5 means the fifth volume value. It can mean a volume value.

delete

6 is a diagram illustrating derivation of a fifth height value according to an embodiment of the present invention.

Referring to FIG. 6 , in order to obtain the volume of the fifth three-dimensional object, which is the overlapping area of the second and fourth three-dimensional objects, the fifth height value must first be derived. Therefore, as shown in FIG. 6, the fifth height value can be derived by simplifying the view from the side of the vehicle, and using Equation 3 as a ratio of lengths.

Looking in more detail, the fifth height value may be derived by Equation 3 below.

[Equation 3]

In this case, H(Height)_5 may mean the fifth height value, H_2 may mean the second height value, and H_4 may mean the fourth height value.

In addition, the processor 110 may derive the average specific gravity of the mixed scrap waste based on the metal type and the metal ratio included in the DB including the specific gravity information for each metal type and the mixed information.

For example, if the included metal types are iron, copper, and zinc, the metal ratio is 5:3:1, and each specific gravity is 3, 2, or 4, the average specific gravity is about 2.76 by the equation below can be derived

Also, the processor 110 may derive the weight information based on the volume information and the average specific gravity.

delete

7 is a diagram illustrating extraction of a small object and a large object according to an embodiment of the present invention.

Referring to FIG. 7 , the processor 110 may classify all objects included in the second image through an artificial intelligence module, set them as small objects, and create outlines surrounding the edges of each small object. At this time, regardless of overlapping, everything that looks like an object can be set as a small object.

In addition, the processor 110 generates a virtual extension line by extending the outline with the same curvature or in the same direction based on the corner, when each small object has a corner on the outline, and the extension lines are mutually related to each other. It is possible to extract other connected small objects, collect other small objects connected through an extension line, and set a large object including the outline of the corresponding small objects and the extension line.

At this time, when objects overlap each other, a corner inevitably occurs at the overlapping portion. Therefore, it can be seen that the corresponding small objects form one object depending on whether the extension lines extending the outline based on the corresponding edge connect to each other. In summary, the corresponding small objects can be seen as one object that is covered by the object in front. Accordingly, an object connected by the extension line may be set as a large object and set as one object for determining the type of metal.

In addition, the processor 110 may set only the corresponding small object as a large object of a small object without other small objects connected through the extension line. This is because there are cases where there are no other objects obscuring the object.

Also, the processor 110 may derive the metal type of the target object based on the color information of the target object. In this regard, it will be described later in detail.

Also, the processor 110 may derive the metal ratio for each metal type based on the area occupied by all large objects in the second image.

In this case, the metal ratio may mean a relative ratio for each metal type of the object included in the second image, and the metal ratio is obtained by deriving the relative ratio based on the number of pixels of the object according to the metal type. can be set to

For example, if the number of pixels of the large object corresponding to iron is 10000, the number of pixels of the large object corresponding to copper is 100, and the number of pixels of the large object corresponding to aluminum is 2000, the metal ratio can be set to 100:1:20 in the order of iron, copper and aluminum.

Also, the processor 110 may generate the mixed information based on the metal type and the metal ratio.

delete

Metals can be distinguished by color. Accordingly, the type of the corresponding metal may be extracted based on the RGB values of the pixels of the object.

8 is a diagram showing an operation outline of a similarity module according to an embodiment of the present invention.

Referring to FIG. 8 , the processor 110 may derive large-object general color information for the corresponding large-object as an average of RGB values of pixels in the area occupied by the small object in the area occupied by the large-object.

In addition, the processor 110 divides each metal into first color information representing RGB values preset to correspond to the general color of the metal and second color information representing RGB values preset to correspond to the corrosion color of the metal. Based on the metal color DB, a first similarity between the large object color information of the large object and the first color information may be derived for each large object through a preset similarity module.

In this case, the metal color DB may define average RGB values for colors in a normal state and colors in a corroded state according to the type of metal.

For example, in the case of iron, it may be gray or black in a normal state, and may be orange or reddish brown in a corroded state. Accordingly, the first color information and the second color information may be set as average values of RGB values in each state.

As another example, copper may be orange or reddish brown in normal state and green in corroded state, and zinc may be silver or gray in normal state and light gray or white in corroded state. can be

At this time, the operation of the similarity module will be described later in more detail.

In addition, the processor 110 may set a plurality of metals whose first similarity exceeds a preset first threshold similarity as first metal candidates. This is because there may be several metals with similar colors in their normal state.

In this case, the first critical similarity may be set as an average of first similarities for all metals.

As described above, since there may be several metals having similar colors in a general state, the metal type of the target object can be more accurately derived through color change due to metal corrosion.

In more detail, the processor 110 may derive the large-object corrosion color information for the corresponding large-object as an average of the RGB values of the pixels of the corroded area derived as the corroded portion among the areas occupied by the small object. At this time, since there is a high possibility that only a portion of the corroded area is corroded, the processor 110 derives a pixel most similar to the general color information of the large object from among the pixels of the large object, and locates the corresponding pixel. It is possible to set the area to be the corrosion area, and the anti-object corrosion color information can be derived based on the RGB values of the corresponding pixels.

In more detail, the processor 110 derives a third similarity between the RGB value of each pixel of the area occupied by the small object and the large object general color information through the similarity module, and among the total third similarity The lowest 3-1 similarity may be derived, and pixels included from the 3-1 similarity to a preset lowest similarity range may be derived as the corrosion area.

In this case, the lowest similarity range may be arbitrarily set, and may be set to, for example, 10%.

In addition, the processor 110 derives a second similarity between the second color information of the metal corresponding to the first metal candidate and the object corrosion color information through the similarity module, and the second similarity is The highest metal can be set as the metal type of the target object.

At this time, the similarity module generates a first vector having R, G, and B values as components based on the input first RGB values, and based on the input second RGB values together with the first RGB values. As a result, a second vector having each value of R, G, and B as a component can be generated, and the degree of similarity can be derived through Equation 4 below.

[Equation 4]

In this case, PS (Pixel Similarity) may mean the similarity, V (Vector) 1_i may mean the i-th component of the first vector, and V2_i may mean the i-th component of the second vector.

Taking the first similarity derivation as an example, the first vector may be V1 = (R1, G1, B1) generated based on the RGB values included in the large object general color information, and the second vector may be the first vector. It may be V2 = (R2, G2, B2) generated based on RGB values for a specific metal included in the color information, and the first similarity may be derived by Equation 1 based on the V1 and the V2. can

In addition, the processor 110 may derive the metal type and metal ratio included in the mixed scrap iron waste based on the above-described configuration, and derive the mixed unit price information based on this, provided that the degree of mixing is If it is very high, it is desirable to reflect this in the transaction cost, since a considerable amount of time and labor must be invested in the sorting operation.

Therefore, the processor 110 determines the amount of the mixed scrap waste based on the classification weight derived based on the degree of agglomeration of each metal type included in the mixed scrap metal waste, the unit price per unit weight preset for each metal type, and the metal ratio. The mixed unit price information may be derived.

delete

9 is a conceptual diagram for deriving classification weights according to an embodiment of the present invention.

Referring to FIG. 9 , the mixed unit price information may be derived by Equation 5 below.

[Equation 5]

At this time, MUP (Mixed Unit Price) means the mixed unit price information, CW (Classification Weight) means the classification weight, and RR (Relative Ratio)_j is the jth metal type included in the mixed scrap iron waste. It means a relative ratio, UP (Unit Price)_j means the unit price per unit weight of the j-th metal type included in the mixed scrap iron waste, and n may mean the number of the metal type.

At this time, the classification weight is derived based on how much each metal type is gathered together, and looking in more detail, the processor 110 divides a lattice-type layer including blocks of a preset size into the second image and the second image. overlap, and the block may be divided into a first block in which the large object occupies more than half of the block and a second block in which the large object does not occupy more than half of the block.

In this case, the size of the block may be arbitrarily set by the user.

In addition, the processor 110 classifies the first block according to the metal type, selects an arbitrary third block from blocks classified according to the metal type for each metal type, and The distance between the center point and the center point of the remaining blocks is extracted, the average block distance value indicating the average value of the distance is extracted, and the smallest value among the average block distance values of the third blocks is extracted as the minimum block distance value of the corresponding metal type. can do.

Through this, it is possible to check where the specific metal is most concentrated in the second image and how much it is spread.

Also, the processor may derive the classification weight based on the minimum block distance value for each metal type and the size of the second image.

Looking in more detail, the classification weight can be derived by Equation 6 below.

[Equation 6]

At this time, CW (Classification Weight) means the classification weight, DI (Distance Index) means a distance index meaning the degree of aggregation of metal types included in the mixed metal waste, and A is the second image It may mean a longer length of the horizontal or vertical length.

In addition, the distance index may be derived by Equation 7 below.

[Equation 7]

At this time, DI (Distance Index) means the distance index, Av_CDis_min_j means the minimum block distance value of the j-th metal type, and n may mean the number of the metal type.

Through this, it is possible to derive mixed unit price information according to the type and ratio of metals included in the mixed scrap waste, and accordingly, to derive transaction cost along with weight information.

delete

10 is a flowchart of a method for predicting the weight of mixed scrap waste based on image analysis according to an embodiment of the present invention.

Referring to FIG. 10, in the method for predicting the weight of mixed scrap waste based on image analysis according to an embodiment of the present invention, the transport vehicle carrying the mixed scrap waste is photographed in a perspective view direction according to the movement of the transport vehicle carrying the mixed scrap metal waste. The first image may be received from a camera that is installed and generates a first image of the mixed scrap waste and the transportation means (S101).

In addition, the method for predicting the weight of mixed scrap metal based on image analysis according to an embodiment of the present invention analyzes the first image to generate mixed information including metal types and metal ratios of metals included in the mixed scrap metal waste. It can (S103).

In addition, the method for predicting the weight of mixed scrap waste based on image analysis according to an embodiment of the present invention may generate weight information of the mixed scrap waste by analyzing the first image (S105).

In addition, the method for predicting the weight of mixed scrap waste based on image analysis according to an embodiment of the present invention may derive the transaction cost of the mixed scrap waste based on the mixed information and the weight information (S107).

delete

In addition, the method for predicting the weight of mixed scrap waste based on image analysis according to an embodiment of the present invention may have the same configuration as the device for predicting the weight of mixed scrap waste based on image analysis disclosed in FIGS. 1 to 9 .

delete

The embodiments described above may be implemented as hardware components, software components, and/or a combination of hardware components and software components. For example, the devices, methods, and components described in the embodiments may include, for example, a processor, a controller, an arithmetic logic unit (ALU), a digital signal processor, a microcomputer, a field programmable gate (FPGA). array), programmable logic units (PLUs), microprocessors, or any other device capable of executing and responding to instructions. A processing device may run an operating system (OS) and one or more software applications running on the operating system. A processing device may also access, store, manipulate, process, and generate data in response to execution of software. For convenience of understanding, there are cases in which one processing device is used, but those skilled in the art will understand that the processing device includes a plurality of processing elements and/or a plurality of types of processing elements. It can be seen that it can include. For example, a processing device may include a plurality of processors or a processor and a controller. Other processing configurations are also possible, such as parallel processors.

The method according to the embodiment may be implemented in the form of program instructions that can be executed through various computer means and recorded on a computer readable medium. The computer readable medium may include program instructions, data files, data structures, etc. alone or in combination. Program commands recorded on the medium may be specially designed and configured for the embodiment or may be known and usable to those skilled in computer software. Examples of computer-readable recording media include magnetic media such as hard disks, floppy disks and magnetic tapes, optical media such as CD-ROMs and DVDs, and magnetic media such as floptical disks. - includes hardware devices specially configured to store and execute program instructions, such as magneto-optical media, and ROM, RAM, flash memory, and the like. Examples of program instructions include high-level language codes that can be executed by a computer using an interpreter, as well as machine language codes such as those produced by a compiler. The hardware devices described above may be configured to operate as one or more software modules to perform the operations of the embodiments, and vice versa.

Software may include a computer program, code, instructions, or a combination of one or more of the foregoing, which configures a processing device to operate as desired or processes independently or collectively. The device can be commanded. Software and/or data may be any tangible machine, component, physical device, virtual equipment, computer storage medium or device, intended to be interpreted by or provide instructions or data to a processing device. , or may be permanently or temporarily embodied in a transmitted signal wave. Software may be distributed on networked computer systems and stored or executed in a distributed manner. Software and data may be stored on one or more computer readable media.

As described above, although the embodiments have been described with limited drawings, those skilled in the art can apply various technical modifications and variations based on the above. For example, the described techniques may be performed in an order different from the method described, and/or components of the described system, structure, device, circuit, etc. may be combined or combined in a different form than the method described, or other components may be used. Or even if it is replaced or substituted by equivalents, appropriate results can be achieved.

Therefore, other implementations, other embodiments, and equivalents of the claims are within the scope of the following claims.

delete

Claims (5)

In electronic devices,
memory; and a processor connected to the memory; including,
The processor:
The first image is obtained from a camera that is installed to photograph the transportation means carrying the mixed scrap waste in a perspective view along the movement of the transportation means loaded with the mixed scrap metal waste, and generates a first image of the mixed scrap metal waste and the transportation means. receive,
Analyzing the first image, generating mixed information including metal types and metal ratios of metals included in the mixed scrap iron waste;
Analyzing the first image to generate weight information of the mixed scrap metal waste;
Deriving a transaction cost of the mixed scrap waste based on the mixed information and the weight information;
The processor:
Through an artificial intelligence module, a second image is generated by extracting only the mixed scrap waste from the first image;
Through an artificial intelligence module, a receiving space accommodating the mixed scrap metal waste is identified in the second image, and among the identified receiving spaces, a rear plate positioned on the rear side of the means of transportation and a space located on the side of the means of transportation are identified. Separate the side plate to
A lower line of the rear plate is set as a first lower line,
Set the lower line of the side plate as a second lower line,
Set the leftmost point in the second image as the first point,
Set the rightmost point in the second image as the second point,
connecting a first vertical line descending vertically from the first point to the first lower line;
connecting a second vertical line descending vertically from the second point to the second lower line;
generating a third image including the second image, the first vertical line, the second vertical line, the first lower line, and the second lower line;
Generating the weight information based on the third image;
The processor:
Based on the third image, volume information of the mixed scrap waste is generated;
Generating weight information of the mixed scrap waste based on a DB including specific gravity information for each metal type, the volume information, and the mixing information;
The processor:
By comparing the lengths of the first vertical line and the second vertical line based on the third image,
If the difference between the first vertical line and the second vertical line is within a preset error range, based on the first volume value of the first solid object of the first shape and the second volume value of the second solid object of the second shape. To generate the volume information as a first type of generating the volume information,
If the difference between the first vertical line and the second vertical line is out of the error range, the second volume value, the third volume value of the third solid object of the first shape, and the fourth volume value of the fourth solid object of the second shape The volume information is generated in a second type for generating the volume information based on a volume value and a fifth volume value of the fifth three-dimensional object of the second shape;
The first type is:
An average of the first vertical line and the second vertical line is set as a first height value,
Set the length of the first lower line as a horizontal value,
Set the length of the second lower line as a vertical value,
The first shape is a rectangular parallelepiped, and the first volume value of the first three-dimensional object is derived based on the first height value, the horizontal value, and the vertical value,
Set the point located at the top of the third image as the top point,
Create a top vertical line that goes down vertically from the top point;
generating a top reference line connecting the first point and the second point;
A length between a central point where the tower vertical line and the tower reference line meet and the top point is set as a second height value;
The second shape is a quadrangular pyramid, and the second volume value of the second three-dimensional object is derived based on the second height value, the horizontal value, and the vertical value,
A type of generating the volume information through Equation 1 below,
[Equation 1]

VI (Volume Information) means the volume information, VV (Value of Volume)_1 means the first volume value, VV_2 means the second volume value,
The second type is:
A line having a short length among the first vertical line and the second vertical line is set as a third vertical line, a length of the third vertical line is set as a third height value, and the third height value, the horizontal value, and the vertical value Deriving the third volume value of the third solid object based on
A length difference between the first vertical line and the second vertical line is set as a fourth height value, and the fourth volume value of the fourth three-dimensional object is derived based on the fourth height value, the horizontal value, and the vertical value. do,
A fifth height value of the fifth three-dimensional object is derived based on the second height value and the fourth height value, and the fifth height value of the fifth three-dimensional object is derived based on the fifth height value, the horizontal value, and the vertical value. 5 derive the volume value,
A type of generating the volume information through Equation 2 below,
[Equation 2]

VI denotes the volume information, VV_2 denotes the second volume value, VV_3 denotes the third volume value, VV_4 denotes the fourth volume value, and VV_5 denotes the fifth volume value. An electronic device, characterized in that means. delete delete delete delete

Images