KR102250864B1 - Method and device for cutting by quantity based on artificial intelligence - Google Patents

Abstract

According to an embodiment, provided is a method for controlling fixed quantity cutting based on artificial intelligence, which is performed by a device. The method comprise the steps of: obtaining, when an object to be cut is scanned by a scanner unit, a 3D image of the object to be cut from the scanner unit; dividing the object to be cut into a plurality of areas based on the 3D image, and calculating a volume for each area of the object to be cut; obtaining, when the weight of the object to be cut is measured by a measurement unit, the weight of the object to be cut from the measurement unit; calculating the weight for each area of the object to be cut based on the volume for each area of the object to be cut and the weight of the object to be cut; outputting a control value for cutting the object to be cut into a fixed quantity unit by applying a type of the object to be cut, the fixed quantity unit and the weight for each area to artificial intelligence; and setting a type of a blade of a cutting unit, a cutting position, and a belt moving speed of a conveyor unit through the control value, and controlling the object to be cut so that the object to be cut is cut into the fixed quantity unit.

Description

Quantitative cutting method and device based on artificial intelligence {METHOD AND DEVICE FOR CUTTING BY QUANTITY BASED ON ARTIFICIAL INTELLIGENCE}

The following examples relate to a technology for controlling quantitative cutting based on artificial intelligence.

Previously, meat such as growth and seasoned meat was managed by agricultural and livestock products, but the Food and Drug Administration, which manages pharmaceuticals, foods, cosmetics, etc., managed instant processed foods and seasoned meats, and managed by the introduction of HACCP (Hazard Analysis and Critical Control Point) certification. Is getting stricter.

For example, when slaughtering a cow, check the mother's ID, sperm ID, the cow's ID, the date the cow was born, the date of slaughter, the weight at the time of slaughter, whether it was castrated, etc., cut the head, forefoot and hind paws, and After peeling the skin, removing the intestines, and hanging the meat from side to side according to a predetermined tool, the process of hanging the meat is carried out. At this time, it is recorded for each part, such as a major classification, a neck core, a sirloin, and a scallop, and the weight for each part may be determined.

Due to the implementation of the history system, weight information such as the total weight of the meat and the left and right weight of each part must be tabulated and reported to the Livestock Product Quality Assessment Service, so it was possible to prevent corruption in the meat distribution process through the checklist. At this time, the retailer is obligated to report information on when it was sold for each part.

Meat grades can be classified into ++1, +1, 1st grade, 2nd grade, 3rd grade, and other grades, and there is a guideline amount for each meat part, and it will be decided by auction. At this time, when determining the meat part grade, the meat part may be graded by judging with only partial complaints such as marbling.

Even if it's the same sirloin, it's tough when it's close to the neck core, and it tastes good when it's close to the tip of the stalk. , For example, there is a tendency to purchase only sirloin. Accordingly, there is a problem in that the remaining parts other than the ribs are inevitably accumulated, and since they are fresh and new, they must be discarded if the stock remains. For example, consumers only look at the backside of the sirloin and judge the grade of the whole sirloin as the masses of the eyes, so they are exposed to uncertainty that may change the actual state of the meat inside.

In particular, in the case of online sales, there is a problem that it is not possible to select each detailed part, because the actual confirmation is not possible, and only good detailed parts can be exposed what is shown in the image. In addition, there is a problem that weight loss occurs, and the weight is not constant for each product to be sold.

On the other hand, if meat (beef, pork, chicken, duck, etc.), aquatic products (fish, shellfish, seaweed, etc.), agricultural products, forest products, etc. are processed and packaged in large quantities, uniform weight (hereinafter referred to as "quantity") Packaging is done in units. And the packaged products are distributed nationwide through marts, traditional markets, home shopping, and internet order delivery. Therefore, if a single cut body (meat, aquatic products, agricultural products, etc.) has different weights and is larger than the amount required for processing, it is necessary to cut it in a fixed amount during processing.

Moreover, in the case of repetitive cutting by dividing a large group to be cut into small and uniform quantities, the cutting operation must be performed quickly and accurately.

However, in the related art, workers cut the group to be cut by their own experience or cut the group to be cut in a standardized manner. And the weight of the cut body was measured in units of 1g to calculate the price. In addition, the repetitive cutting operation had to rely entirely on skilled workers.

Therefore, it was very difficult to automatically cut the cut body continuously in a quantitative unit. For example, for larger cuts, the quantity can be adjusted manually, but for smaller cuts, it has to be discarded or used for other purposes. In particular, since the condition of the cut group, which is worn every day for cutting work, varies according to the season, the place of origin, and the form of aquaculture, research and development on a technology capable of quantitatively cutting in large quantities is required.

Korean Patent Publication No. 10-2148940 (2020.08.28)

According to an embodiment, the group to be cut is divided into a plurality of areas based on the 3D image of the group to be cut, the volume of the group to be cut is calculated, and the group to be cut is based on the volume of the group to be cut and the weight of the group to be cut. Calculate the weight for each area of the target, apply the type of the entity to be cut, the unit of quantification, and the weight by area to artificial intelligence, output a control value for cutting the entity to be cut in a quantitative unit, and output the control value for cutting the entity to be cut in a quantitative unit. It is an object of the present invention to provide an artificial intelligence-based quantitative cutting method and apparatus for controlling a cutting position and a belt movement speed of a conveyor unit to be cut in a quantitative unit.

The object of the present invention is not limited to the above-mentioned object, and other objects not mentioned will be clearly understood from the following description.

According to an embodiment, in a method for controlling quantitative cutting based on artificial intelligence, performed by an apparatus, when a cut entity is scanned by a scanner unit, a 3D image of the cut entity is obtained from the scanner unit. step; Dividing the group to be cut into a plurality of areas based on the 3D image and calculating a volume for each area of the group to be cut; When the weight of the to-be-cut entity is measured by a measuring unit, obtaining a weight of the to-be-cut entity from the measurement unit; Calculating a weight for each area of the group to be cut based on the volume of the group to be cut and the weight of the group to be cut; Outputting a control value for cutting the cut entity into a quantitative unit by applying the type, a unit of quantification, and the weight of each area to the artificial intelligence; And setting a blade type of a cutting unit, a cutting position, and a belt movement speed of a conveyor unit through the control value, and controlling the cutting unit to be cut in the quantitative unit.

Calculating the volume of the group to be cut for each area includes calculating the volume of the group to be cut by summing the volume for each area of the group to be cut, and calculating the weight for each area of the group to be cut includes: Calculating the density of the to-be-cut body by dividing the weight of the to-be-cut body by the volume of the to-be-cut body; Determining which part of the group to be cut is a first area, which is one of the plurality of areas of the group to be cut, based on the 3D image; If the first region is identified as a first region, obtaining a density of the first region from a database; When the difference between the density of the cut body and the density of the first portion is identified within a predefined error range, the weight of the first region is calculated by multiplying the volume of the first region and the density of the cut body. step; And when it is determined that the difference between the density of the to-be-cut entity and the density of the first region is outside a predefined error range, the value obtained by multiplying the volume of the first region and the density of the first region, It may include the step of calculating the weight.

In the artificial intelligence-based quantitative cutting method, when the cutting body is cut through a blade of the cutting unit and separated into a plurality of cutting bodies, a first image of the cut surface of the first cutting body, which is one of the plurality of cutting bodies, is Obtaining; Classifying the first cut body into a plurality of zones according to the internal state of the first cut body based on the first image, and predicting and calculating a density for each zone of the first cut body; Obtaining a price of the first state from a database when a first region, which is one of the plurality of regions of the first cutting body, is identified as a first state through density; Calculating the weight in the first state by multiplying the weight of the first cut body and the ratio of the area occupied by the first area in the cut surface of the first cut body; Calculating a price of the first zone by multiplying the weight of the first state and the price of the first state; And calculating a price of the first cutting body by summing the price calculated for each area of the first cutting body.

The controlling of the cut body to be cut into the unit of quantification includes a value obtained by subtracting 1 from an integer value obtained by increasing a value obtained by dividing the weight of the body to be cut by the unit of quantification, and cutting the body to be cut into the unit of measure. Setting the required number of blades of the cutting unit; And setting a cutting position in which the blades corresponding to the number of blades of the cutting unit protrude based on the weight information for each region.

The artificial intelligence-based quantitative cutting method, when the practice mode for manually cutting the cut body is selected, generates contents of cutting exercise for cutting the cut body into the quantitative unit using the 3D image, and the Controlling the cutting practice content to be displayed on the display unit; When an input signal through a user's touch is detected on the display unit, detecting a position of the user's hand through the input signal, tracking the movement of the user's hand, and analyzing the movement of the user; And evaluating the cutting ability of the user based on the analysis result of the moving movement line, wherein the controlling to display the cutting practice content comprises: when a position of the user's hand is detected at a point indicated as a starting point, Controlling the target point to be displayed; Controlling a guide mark for guiding a movement line moving from the starting point to the target point to be displayed in an area adjacent to the position of the user's hand; Controlling the guide mark to move and display according to the movement of the hand position; And obtaining a first route through the movement of the user's hand moved in a first period, and predicting a first position of the guide mark to be displayed in a second period after the first period based on the first route. , Checking the second position of the guide mark to be displayed in the second period through a second path set in the cutting practice content, and based on the vector difference value between the first position and the second position, of the guide mark Setting a size and controlling the guide mark to be displayed larger as the difference value increases, and the controlling to display the target point includes moving from the starting point through the first point and the second point to the target point. Controlling the first point to be further displayed while the starting point is highlighted when it is detected that the user's hand is located at the starting point while the moving line is set as the cutting practice content; When it is detected that the user's hand position is out of the starting point, controlling the second point to be further displayed while the first point is highlighted; Controlling the starting point not to be displayed when the position of the user's hand is moved and the time out of the starting point is determined to be greater than or equal to a reference time; Predicting a moving time to the first point based on the first route, and controlling the second point to be highlighted before the user's hand moves to the first point through the predicted moving time ; If it is sensed that the user's hand is located at the first point, controlling the target point to be further displayed; Controlling the first point not to be displayed when the user's hand position is moved and the time out of the first point is determined to be greater than or equal to a reference time; And predicting a movement time to the second point based on the first route, and controlling the target point to be highlighted before the user's hand moves to the second point through the predicted movement time. Can include.

According to an embodiment, the group to be cut is divided into a plurality of areas based on the 3D image of the group to be cut, the volume of the group to be cut is calculated, and the group to be cut is based on the volume of the group to be cut and the weight of the group to be cut. Calculate the weight for each area of the target, apply the type of the entity to be cut, the unit of quantification, and the weight by area to artificial intelligence, output a control value for cutting the entity to be cut in a quantitative unit, and output the control value for cutting the entity to be cut in a quantitative unit. By setting the cutting position and the belt movement speed of the conveyor unit, and controlling the cutting unit to be cut in units of quantity, the unit to be cut can be repeatedly cut in units of quantity, so that the processing quality of the unit to be cut is standardized, and the condition of the unit to be cut and Regardless, there is an effect that quantitative cutting in large quantities is possible.

Meanwhile, the effects according to the embodiments are not limited to those mentioned above, and other effects not mentioned will be clearly understood by those of ordinary skill in the relevant technical field from the following description.

1 is a diagram schematically showing the configuration of an apparatus for quantitatively cutting an entity to be cut according to an embodiment.
FIG. 2 is a diagram for explaining a process of controlling a cut entity to be cut in a quantitative unit according to an exemplary embodiment.
3 is a view showing a side surface of an entity to be cut divided into a plurality of areas according to an exemplary embodiment.
4 is a view for explaining a process of calculating a weight according to the density of each part according to an embodiment.
5 is a view for explaining a process of calculating a price of a cut body according to an embodiment.
6 is a view showing a cut surface of a first group to be cut divided into a plurality of zones according to an embodiment.
7 is a diagram for explaining learning of an artificial neural network according to an embodiment.
8 is a diagram illustrating a process of evaluating a user's cutting ability through cutting practice content according to an embodiment.
9 is a diagram for describing a display of a guide mark according to an exemplary embodiment.
10 is a diagram for describing a display of a starting point, a first point, a second point, and a target point according to an exemplary embodiment.

Hereinafter, embodiments will be described in detail with reference to the accompanying drawings. However, since various changes may be made to the embodiments, the scope of the patent application is not limited or limited by these embodiments. It is to be understood that all changes, equivalents, or substitutes to the embodiments are included in the scope of the rights.

Specific structural or functional descriptions of the embodiments are disclosed for the purpose of illustration only, and may be changed and implemented in various forms. Accordingly, the embodiments are not limited to a specific disclosure form, and the scope of the present specification includes changes, equivalents, or substitutes included in the technical idea.

Although terms such as first or second may be used to describe various components, these terms should be interpreted only for the purpose of distinguishing one component from other components. For example, a first component may be referred to as a second component, and similarly, a second component may be referred to as a first component.

When a component is referred to as being "connected" to another component, it is to be understood that it may be directly connected or connected to the other component, but other components may exist in the middle.

The terms used in the examples are used for illustrative purposes only and should not be construed as limiting. Singular expressions include plural expressions unless the context clearly indicates otherwise. In the present specification, terms such as "comprise" or "have" are intended to designate the presence of features, numbers, steps, actions, components, parts, or combinations thereof described in the specification, but one or more other features. It is to be understood that the presence or addition of elements or numbers, steps, actions, components, parts, or combinations thereof does not preclude in advance.

Unless otherwise defined, all terms used herein, including technical or scientific terms, have the same meaning as commonly understood by one of ordinary skill in the art to which the embodiment belongs. Terms as defined in a commonly used dictionary should be interpreted as having a meaning consistent with the meaning in the context of the related technology, and should not be interpreted as an ideal or excessively formal meaning unless explicitly defined in the present application. Does not.

In addition, in the description with reference to the accompanying drawings, the same reference numerals are assigned to the same components regardless of the reference numerals, and redundant descriptions thereof will be omitted. In describing the embodiments, when it is determined that a detailed description of related known technologies may unnecessarily obscure the subject matter of the embodiments, the detailed description thereof will be omitted.

The embodiments may be implemented in various types of products such as a personal computer, a laptop computer, a tablet computer, a smart phone, a television, a smart home appliance, an intelligent vehicle, a kiosk, and a wearable device.

1 is a diagram schematically showing the configuration of an apparatus for quantitatively cutting an entity to be cut according to an embodiment.

As shown in FIG. 1, the apparatus 100 may include a scanner unit 110, a measurement unit 120, a conveyor unit 130, a cutting unit 140, a display unit 150, and a control unit 160. I can.

First, the device 100 may be configured to perform all or part of a calculation function, a storage/referencing function, an input/output function, and a control function of a conventional computer. The device 100 may include at least one artificial neural network that performs an inference function. The device 100 may be configured to communicate with external devices by wire or wirelessly.

The scanner unit 110 may scan the group to be cut and generate a 3D image of the group to be cut. The scanner unit 110 may be configured with various devices that generate a 3D image of an object to be cut, such as a 3D scanner, a plurality of cameras, a laser, and a lidar. Here, the cut body is an atypical object, and may include various objects for cutting such as livestock products, meat, and fish.

The scanner unit 110 is installed on the upper side from the front end of the conveyor unit 130, and when the to be cut is placed on the front end of the conveyor unit 130, the scanner unit 110 scans the to be cut and You can create a 3D image for

The scanner unit 110 may generate a 3D image by restoring the appearance of the group to be cut in 3D, and the volume and weight of the group to be cut may be calculated through the 3D image, and the weight for each area of the group to be cut may be calculated. .

The measuring unit 120 is installed from the front end to the lower side of the conveyor unit 130, and when the to-be-cut entity is placed on the front end of the conveyor unit 130, the measuring unit 120 is to measure the weight of the to-be-cut entity. I can. The measuring unit 120 may be configured as an electronic scale or a digital scale.

The conveyor unit 130 is a means for moving the cut unit, and when the belt of the conveyor unit 130 moves, the cut unit placed on the conveyor unit 130 may move along the belt of the conveyor unit 130.

The cutting unit 140 is installed at the rear end of the conveyor unit 130, and when the to be cut is moved to the rear end along the belt of the conveyor unit 130, one or more blades protrude upward or downward so that the to be cut is cut. I can make it. When the to-be-cut entity passes through the rear end of the conveyor unit 130, it may be separated into a plurality of cut bodies by the blade of the cutting unit 140.

The display unit 150 may be configured as a touchable screen, and may display a control operation, input/output, etc. of the device 100 or receive selection information from a user.

The control unit 160 controls the scanner unit 110, the measurement unit 120, the conveyor unit 130, the cutting unit 140, and the display unit 150 included in the device 100 so that each operation is normally performed. I can.

According to an embodiment, the device 100 may include a database or may be connected to a database. The database may be in a state in which various information necessary for quantitative cutting, such as density information for each part and price information for each state, is stored.

The controller 160 may perform at least one method described later with reference to FIGS. 2 to 10.

FIG. 2 is a diagram for explaining a process of controlling a cut entity to be cut in a quantitative unit according to an exemplary embodiment.

Referring to FIG. 2, first, in step S201, when the to-be-divided entity is scanned by the scanner unit 110, the apparatus 100 may obtain a 3D image of the to-be-divided entity from the scanner unit 110.

Specifically, when the to-be-cut entity is placed on the front end of the conveyor unit 130, the scanner unit 110 may scan the to-be-cut entity to generate a 3D image of the cut-off entity, and the device 100 3D images can be acquired. Here, the scanner unit 110 may obtain a 3D image of the group to be cut by scanning the volume of the group to be cut through a plurality of cameras. For example, the volume of the group to be cut may be scanned through cameras installed on the left side, the right side, the upper side, and the front side of the group to be cut. The camera installed on the front of the group to be cut may be hidden by moving downwards or upwards so as not to obstruct the movement of the group to be cut after shooting.

In step S202, the apparatus 100 may divide the group to be cut into a plurality of areas based on the 3D image.

For example, the device 100 may classify the group to be cut into a first area, a second area, a third area, a fourth area, etc. by dividing the group to be cut at regular intervals.

In step S203, the apparatus 100 may calculate the volume for each area of the to-be-divided entity based on the 3D image.

For example, when the to-be-cut entity is divided into a first area, a second area, a third area, and a fourth area, the device 100 sets the volume of the first area to 10 cm ³ and the volume of the second area to 20 cm. ³ , the volume of the third area may be calculated ^{as 30 cm 3} , and the volume of the fourth area ^{may be calculated as 40 cm 3.} At this time, the device 100 may calculate a volume ratio that each region occupies in the entire region of the to be cut, and the volume ratio occupied by the first region is 10%, and the volume ratio occupied by the second region is 20%. , The volume ratio occupied by the third region may be calculated as 30%, and the volume ratio occupied by the fourth region may be calculated as 40%. That is, even if the to-be-divided entity is divided into a plurality of regions by being divided at equal intervals, since the to-be-divided entity is an atypical object, the volume of each region may be different.

In step S204, when the weight of the object to be cut is measured by the measuring unit 120, the device 100 may obtain the weight of the object to be cut from the measuring unit 120.

Specifically, when the to-be-divided entity is placed on the front end of the conveyor unit 130, the measurement unit 120 may measure the weight of the to-be-cut entity, and the device 100 may obtain the weight of the to-be-cut entity.

In step S205, the apparatus 100 may calculate the weight for each area of the group to be cut based on the volume of the group to be cut and the weight of the group to be cut.

For example, the volume of the first area is 10 cm ³ , the volume of the second area is 20 cm ³ , the volume of the third area is 30 cm ³ , the volume of the fourth area is 40 cm ^3, and the weight of the cut body is If 100g is determined, the device 100 may calculate the weight of the first area as 10g, the weight of the second area as 20g, the weight of the third area as 30g, and the weight of the fourth area as 40g. That is, since the to-be-cut entity is an atypical object, the volume of each region is different, so the weight of each region may be calculated according to the volume of each region.

The device 100 may output a control value for cutting the cut entity into a quantitative unit by applying information on the type, a unit of quantity, and a weight of each area to the artificial intelligence. Here, the control value may mean a control value for setting a blade type of the cutting unit 140, a cutting position, and a belt movement speed of the conveyor unit 130. The type and unit of quantification of the group to be cut may be set in advance with information input through the display unit 150, and the unit of quantification may be set to a different value according to the type of the group to be cut.

Specifically, the device 100 may include a first artificial intelligence (first artificial neural network) learned in advance. The first artificial neural network may be trained to infer an optimal method of cutting the afflicted entity according to the unit of quantification by receiving input of the type of the entity to be cut, the unit of quantification, and the weight information for each area of the entity to be cut.

The first artificial neural network outputs the blade type, cutting position and belt movement speed in an optimal way to cut the cut group according to the unit by quantification by inferring the type of the group to be cut, the unit of quantification, and the weight of each area of the group to be cut. can do. That is, the first artificial neural network may be trained to infer an optimal method for cutting a cut entity into a quantitative unit based on the type of the cut entity, a quantitative unit, and a weight for each area of the cut entity. The specific learning operation of the first artificial neural network will be described later with reference to FIG. 7.

In step S206, the device 100 may obtain a control value for cutting the to-be-divided entity into quantitative units through the output of the first artificial intelligence.

For example, if the type of the group to be cut is meat, information on the type of blade for meat is output, information on the cutting position of the group to be cut is output according to the unit of quantity and weight by area, and according to the material of the group to be cut. Information on the belt movement speed may be output, and the apparatus 100 may obtain a control value including information on the blade type, the cutting position, and the belt movement speed.

In step S207, the device 100 may control the cutting unit to be cut in a quantitative unit by setting the blade type of the cutting unit 140, the cutting position, and the belt movement speed of the conveyor unit 130 through the control value. . In this case, the to-be-cut entity may move to the cutting unit 140 through the belt of the conveyor unit 130 and may be cut in a quantitative unit while passing through the cutting unit 140. The conveyor unit 130 may drive the belt to move according to the belt movement speed set through the control value, and the cutting unit 140 selects a blade corresponding to the blade type set through the control value, and the selected blades are controlled. It can be processed to protrude from the cutting position set through the value.

According to an embodiment, the apparatus 100 calculates a value obtained by dividing the weight of the object to be cut by a quantitative unit, calculates an integer value obtained by raising the divided value, and calculates a value obtained by subtracting 1 from the rounded integer value, It can be set to the number of blades of the cutting unit 140 required to cut the cut unit by a quantitative unit.

Table 1 shows the number of blades calculated according to the weight and quantification unit of the cut body according to an embodiment.

The weight of the group to be cut Unit of quantification Divided Integer value rounded up Number of blades 100g 100g One One 0 110g 100g 1.1 2 One 120g 100g 1.2 2 One 130g 100g 1.3 2 One 140g 100g 1.4 2 One 150g 100g 1.5 2 One 160g 100g 1.6 2 One 170g 100g 1.7 2 One 180g 100g 1.8 2 One 190g 100g 1.9 2 One 200g 100g 2.0 2 One 210g 100g 2.1 3 2

Referring to Table 1, when the weight of the cut entity is 100g and the quantitative unit is 100g, since cutting of the cut entity is not required, the number of blades may be set to 0. When the weight of the cut body is more than 100g and 200g or less, in order to separate the cut body into two cut bodies, the number of blades may be set to one.

3 is a view showing a side surface of an entity to be cut divided into a plurality of areas according to an exemplary embodiment.

As shown in FIG. 3, the group to be cut 200 may be classified into a first area, a second area, a third area, and a fourth area. Although the width, which is the distance between the regions, is the same, the volume of each region may be different because the to-be-cut body 200 is an atypical object.

For example, the volume of the first area is 50 cm ³ , the volume of the second area is 25 cm ³ , the volume of the third area is 25 cm ³ , the volume of the fourth area is 50 cm ³ , and the cut body 200 When the weight of is determined to be 300g, the device 100 may calculate the weight of the first area as 100g, the weight of the second area as 50g, the weight of the third area as 50g, and the weight of the fourth area as 100g.

According to an embodiment, when the number of blades for cutting the cutting unit 200 in a quantitative unit is set, the device 100 includes blades corresponding to the number of blades of the cutting unit 140 based on the weight information for each area. You can set the protruding cutting position.

For example, when the weight of the cut body 200 is 300 g and the quantitative unit is 100 g, the number of blades may be set to two, and the cut body 200 is a first area, a second area, a third area, If it is divided into a fourth area, and the weight of the first area is 100g, the weight of the second area is 50g, the weight of the third area is 50g, and the weight of the fourth area is 50g, the first area and the second area The cutting position may be set so that the first blade protrudes from the boundary line of, and the cutting position may be set so that the second blade protrudes from the boundary line of the third region and the fourth region. Through this, the first region and the remaining region are cut using the first blade, and the second region, the third region, and the fourth region may be cut using the second blade, so that the first region corresponding to the first region The weight of the cut body is 100g, the weight of the second cut body corresponding to the second area and the third area is 100g, and the weight of the third cut body corresponding to the third area is 100g, and the cut body 200 is a quantitative unit. According to 100 g of phosphorus, it can be cut into three cut bodies.

According to an embodiment, when the to-be-cut body 200 is cut into three cut bodies, the weight of the third cut body may be smaller than the unit of quantification, and the difference between the weight of the third cut body and the unit of quantification increases the error range. When it is determined that the device is out of the way, the device 100 may control the display unit 150 to display a notification message indicating that the third cut body is to be discarded.

For example, if the cut body is 230 g and the quantification unit is 100 g, the weight of the first cut body may be 100 g, the weight of the second cut body 100 g, and the weight of the third cut body may be cut into 30 g. , The weight of the third cut body is 30 g, which is out of the error range from the quantification unit, and a discard notification message for the third cut body may be displayed.

According to an embodiment, the device 100 controls the number of blades protruding upward or downward from the cutting part 140 and the cutting position, and the blades included in the cutting part 140 are in a fixed state, or the device It may be in a state in which it can move under the control of (100).

Specifically, when the blades included in the cutting unit 140 are fixed, the cutting unit 140 includes a plurality of blades, and the plurality of blades are hidden so as not to protrude from the belt of the conveyor unit 130, Only blades selected according to the number of blades and the cutting position may protrude in the belt direction of the conveyor unit 130. When the cutting unit 140 is installed at the upper end of the conveyor unit 130, the blades selected according to the number of blades and the cutting position protrude downward, and the cutting unit 140 is at the lower end of the conveyor unit 130. When installed, the blades selected according to the number of blades and the cutting position may protrude upward.

For example, the first blade, the second blade, the third blade, the fourth blade, the fifth blade, the sixth blade, the seventh blade, the eighth blade, the ninth blade from the left to the right of the cutting part 140 And when the tenth blade is located at 1cm intervals, the number of blades required to cut the unit to be cut in a quantitative unit is set to two, and when the cutting positions at which the two blades protrude are set to 3cm from the left and 7cm from the left, Only the third blade and the seventh blade may protrude in the belt direction of the conveyor unit 130.

When the blades included in the cutting part 140 are in a movable state, the device 100 may control the blades corresponding to the number of blades to move to each cutting position and protrude.

For example, if the number of blades required to cut a single piece to be cut in a quantitative unit is set to two, and the cutting positions at which the two blades protrude are set to 3 cm from the left and 7 cm from the left, the device 100 It is possible to control so that it moves to a cutting position corresponding to 3cm from the left and protrudes, and the second blade moves to a cutting position corresponding to 7cm from the left and protrudes.

4 is a view for explaining a process of calculating a weight according to the density of each part according to an embodiment.

Referring to FIG. 4, first, in step S401, the apparatus 100 may obtain a 3D image of a to-be-cut entity from the scanner unit 110.

In step S402, the apparatus 100 may divide the group to be cut into a plurality of areas based on the 3D image.

In step S403, the device 100 may calculate the volume for each area of the to-be-cut entity based on the 3D image.

In step S404, the apparatus 100 may calculate the volume of the group to be cut by summing the volume of each area of the group to be cut.

In step S405, the device 100 may obtain the weight of the to-be-cut entity from the measurement unit 120.

In step S406, the apparatus 100 may calculate the density of the to-be-cut entity by dividing the weight of the to-be-cut entity by the volume of the to-be-cut entity.

In step S407, the device 100 may determine which part of the first region, which is one of the plurality of regions of the to-be-cut entity, is. In this case, the device 100 may check whether the first region corresponds to which part of the object to be cut is based on the 3D image.

In step S408, when the first region is identified as the first region, the device 100 may obtain the density of the first region from the database. To this end, density information for each region is stored in the database. For example, when the first region is identified as a chest region, density information of the chest region may be obtained through the database.

In step S409, the apparatus 100 may determine whether the difference between the density of the object to be cut and the density of the first portion is within a predefined error range.

If the difference between the density of the cut body and the density of the first part is confirmed to be within a predefined error range in step S409, in step S410, the device 100 multiplies the volume of the first area and the density of the cut body, You can calculate the weight of one area.

For example, when the density of the cut body is 10 g/cm ³ , the density of the first region is 9 g/cm ³ , and the volume of the first region is 10 cm ³ , the device 100 The difference between the densities of the regions is identified within the error range, and 100 g, which is a product of the volume of the first region and the density of the object to be cut, may be calculated as the weight of the first region.

If it is found that the difference between the density of the object to be cut and the density of the first region is outside the predefined error range in step S409, in step S411, the device 100 multiplies the volume of the first region and the density of the first region. As a result, the weight of the first area can be calculated.

For example, when the density of the cut entity is 10 g/cm ³ , the density of the first region is 7 g/cm ³ , and the volume of the first region is 10 cm ³ , the device 100 It is confirmed that the difference between the densities of the regions is out of the error range, and thus 70 g, which is a product of the volume of the first region and the density of the first region, may be calculated as the weight of the first region.

According to one embodiment, since the density may be different depending on the portion of the to-be-cut body, a different weight value may be calculated even with the same volume through this. That is, instead of assuming that the whole to be cut is of the same density, the to be cut is not cut according to the weight of each region, but the density is predicted for each region, and the cut is cut according to the weight of each region in which the difference in density is reflected. In predicting the density of each region to be cut, various methods such as a density measuring device and a plurality of cameras may be employed.

5 is a view for explaining a process of calculating a price of a cut body according to an embodiment.

Referring to FIG. 5, first, in step S501, the device 100 sets the blade type of the cutting unit 140, the cutting position, and the belt movement speed of the conveyor unit 130 through a control value, so that the to-be-cut entity is quantified. It can be controlled to be cut in units. In this case, the to-be-cut body may pass through the cutting part 140 and be cut through the blade of the cutting part 140 to be separated into a plurality of cut bodies.

In step S502, the apparatus 100 may acquire a first image of a cut surface of the first cut body, which is one of the plurality of cut bodies.

For example, the device 100 may acquire a first image by photographing a cut surface with a camera when the to-be-cut entity is cut. To this end, a camera is installed in the cutting unit 140, and when the first cutting body is separated by cutting of the cutting unit, physical pressure may be applied so that the cutting surface of the first cutting body is positioned in the camera direction.

In step S503, the apparatus 100 may classify the first cut body into a plurality of zones according to the internal state of the first cut body based on the first image.

For example, according to the internal condition of the first cutting body, the first cutting body can be classified into a first zone with a quality of 1, a second zone with a quality of 2, a third zone with a quality of 3, etc. .

In step S504, the device 100 may predict and calculate the density for each area of the first cut body.

For example, when the first zone is identified as the first class, the device 100 may predict and calculate the density of the first zone through the first class, and the higher the quality class, the higher the density may be calculated. have.

In step S505, the device 100 may determine which state the first zone is through the density of the first zone, which is any one of the plurality of zones of the first cutting body. To this end, density information for each grade is stored in the database, and when the first area is identified as the first grade, the first grade density information may be obtained through the database.

In step S506, when the first area is identified as the first state through the density, the device 100 may obtain the price of the first state from the database. To this end, price information for each state is stored in the database, and when the first region is determined to be in the first state, price information in the first state may be obtained through the database.

For example, when the first zone is of the first class, the first zone may be identified as the first state through density, and the device 100 may obtain what the price of the first state is according to the unit weight.

In step S507, the apparatus 100 may calculate the weight in the first state by multiplying the weight of the first cutting body and the area ratio occupied by the first area in the cut surface of the first cutting body.

For example, when the weight of the first cut body is 100 g and the area ratio occupied by the first area in the cut surface of the first cut body is 30%, the weight in the first state may be calculated as 30 g.

In step S508, the device 100 may calculate the price of the first zone by multiplying the weight of the first state and the price of the first state.

For example, if the weight of the first state is 30g and the price of the first state is 10 won per 1g, the price of the first zone may be calculated as 300 won.

By repeating steps S505 to S508, prices for the remaining areas of the first cut body other than the first area may be calculated.

In step S509, the apparatus 100 may calculate the price of the first cutting body by summing the price calculated for each area of the first cutting body.

6 is a view showing a cut surface of a first group to be cut divided into a plurality of zones according to an embodiment.

As shown in FIG. 6, the first cutting body 300 may be classified into a first zone, a second zone, and a third zone according to the internal condition, and the density may be predicted and calculated for each zone. have.

Specifically, through the analysis result of the first image of the cut surface of the first cutting body 300, the first area is classified as the area corresponding to the first state, and the second area is divided into the area corresponding to the second state. Classification, and the third zone can be classified as a zone corresponding to the third state.

For example, the first state may be a first grade state including marbling, a second state may be a second grade state including general meat, and the third state may be a third grade state including muscle fat.

The device 100 may calculate the weight of each state by multiplying the weight of the first cutting body 300 and the area ratio occupied by each area in the cut surface of the first cutting body 300.

For example, the weight of the first cutting body 300 is 300 g, the area ratio occupied by the first area in the cut surface of the first cutting body 300 is 20%, and the area ratio occupied by the second area is 70% , When the area ratio occupied by the third zone is 10%, the weight in the first state may be calculated as 60 g, the weight in the second state may be calculated as 210 g, and the weight in the third state may be calculated as 30 g.

The device 100 may calculate the price of the first cutting body 300 according to the price determined for each state.

For example, the weight in the first state is 60g, the weight in the second state is 210g, the weight in the third state is 30g, the price in the first state is 12 won per 1g, and the price in the second state is 10 won per 1g. , If the price in the third state is 8 won per 1g, the price of the first zone is 720 won, the price of the second zone is 2100 won, the price of the third zone is 240 won, and the first cutting body 300 The price of can be calculated at 3060 won.

That is, different prices are applied according to the state of the first cutting body 300, and the price of the first cutting body 300 is calculated by summing them, so that the first cutting body 300 includes a large number of high-quality parts. The higher the price is, the higher the price is, and the lower the number of parts with poor quality is included in the first cutting body 300, the lower the price may be calculated.

7 is a diagram for explaining learning of an artificial neural network according to an embodiment.

The artificial neural network may be a component included in the device 100 and may be learned through the device 100 or a separate learning device.

The first artificial neural network outputs the blade type, cutting position and belt movement speed in an optimal way to cut the cut group according to the unit by quantification by inferring the type of the group to be cut, the unit of quantification, and the weight of each area of the group to be cut. can do.

Hereinafter, a process of learning an artificial neural network through a learning device will be described.

First, in step S701, the training device may acquire training data and a label.

In order to learn the first artificial neural network, the learning device may acquire a data set including information on the type of each member to be cut, a unit of quantification, and a weight for each area of the group to be cut, as each training data. In addition, the learning apparatus may obtain a result obtained by actually cutting the group to be cut into quantitative units according to the type of the group to be cut, the unit of quantification and the weight of each area of the group to be cut, as a label corresponding to each training data.

In step S702, the learning device may generate an input of an artificial neural network from the training data.

The learning apparatus may use the training data as an input of the artificial neural network, or may generate an input of the artificial neural network after a normal process of removing unnecessary information from each training data.

In step S703, the learning device may apply the input to the artificial neural network.

The artificial neural network included in the device 100 may be an artificial neural network that is learned according to supervised learning. The artificial neural network may be a convolutional neural network (CNN) or a recurrent neural network (RNN) structure suitable for training through supervised learning.

In step S704, the learning device may obtain an output from the artificial neural network.

The output of the first artificial neural network may be an inference about an optimal method for cutting a cut entity into a quantitative unit through a type of a cut entity, a unit of quantification, and a weight for each area of the cut entity. Specifically, the first artificial neural network learns the pattern according to the change in the type of the group to be cut, the unit of quantity, and the weight of each area of the group to be cut, and outputs a control value for the optimal method of cutting the group to be cut into units of quantity. I can. In this case, the first artificial neural network may output a control value that helps to set the blade type, the cutting position, and the belt movement speed.

In step S705, the learning device may compare the output and the label. The process of comparing the output of the artificial neural network corresponding to the inference and the label corresponding to the correct answer may be performed by calculating a loss function. As the loss function, a known mean squared error (MSE), cross entropy error (CEE), or the like may be used. However, the present invention is not limited thereto, and loss functions used in various artificial neural network models may be used as long as the deviation, error, or difference between the output of the artificial neural network and the label can be measured.

In step S706, the learning device may optimize the artificial neural network based on the comparison value. The learning device can gradually match the output of the artificial neural network corresponding to the inference and the label corresponding to the correct answer by updating the weight of the nodes of the artificial neural network so that the comparison value becomes smaller and smaller. Can be optimized to output an inference close to the correct answer. Specifically, the learning apparatus may optimize the artificial neural network by repeating the process of resetting the weight of the artificial neural network so that the loss function corresponding to the comparison value approaches the estimated value of the minimum value. For the optimization of artificial neural networks, known backpropagation algorithms, stochastic gradient descent, and the like can be used. However, the present invention is not limited thereto, and a weight optimization algorithm used in various neural network models may be used.

The learning device can train an artificial neural network by repeating this process.

Through this, the first artificial neural network may be trained to output a control value capable of more accurately cutting the cut body in a quantitative unit.

8 is a diagram illustrating a process of evaluating a user's cutting ability through cutting practice content according to an embodiment.

Referring to FIG. 8, first, in step S801, when the exercise mode for manually cutting the cut entity is selected, the apparatus 100 exercises cutting for cutting the cut entity in quantitative units using a 3D image of the cut entity. Content can be generated, and the cutting practice content can be controlled to be displayed on the display unit 150.

The display unit 150 may be implemented as a touch panel in which a user's hand is directly touched, displays various contents according to a practice program as an image, and detects an input signal through a user's touch, and can detect a user's touch. The user's hand position can be detected through an input signal through.

In step S802, when an input signal through a user's touch is detected on the display unit 150, the device 100 detects the position of the user's hand through the input signal, and tracks the user's hand movement to track the movement of the user. Can be analyzed.

In step S803, the device 100 may evaluate the user's cutting ability based on the analysis result of the moving movement line.

In displaying cutting practice content, when the user's hand position is detected at a point indicated as a starting point, the device 100 may control the target point to be displayed, and the speed and vibration of the movement of the user moving from the starting point to the target point , Accuracy and vector values can be analyzed.

The device 100 evaluates whether there is a shaking or tilting phenomenon in a specific direction based on the analysis result of the moving movement line, and checks the reference range of the force for each muscle required to move to the target point, and the force is within the reference range. The entered muscles are classified as normal regions, and the muscles with less or more force outside the reference range are classified as problem regions, so that the user's cutting ability can be evaluated.

9 is a diagram for describing a display of a guide mark according to an exemplary embodiment.

Referring to FIG. 9, the device 100 may control cutting practice content moving from the starting point 401 to the target point 402 to be displayed on the display unit 150.

The device 100 may control the target point 402 to be displayed when the user's hand position is detected at the point indicated as the starting point 401.

The device 100 may recognize the user's hand position 403, and a guide mark 404 for guiding the movement line moving from the starting point 401 to the target point 402 is adjacent to the user's hand position 403. It can be controlled to be displayed in the area.

For example, the device 100 may control the guide mark 404 to be displayed on the top of the hand position 403.

The device 100 may control the guide mark 404 to move and display according to the movement of the hand position 403.

The device 100 may acquire the first route 411 through the movement of the user's hand who moved in the first period, and based on the first route 411, the guide to be displayed in the second period after the first period. The first location 412, which is the location of the marker 404, can be predicted.

The device 100 may check the second position 422 that is the position of the guide mark 404 to be displayed in the second period through the second path 421 set in the cutting practice content.

The device 100 sets the size of the guide mark 404 based on the vector difference value between the first position 412 and the second position 422 so that the larger the difference value, the larger the guide mark 404 is displayed. Can be controlled.

As shown in (a) of FIG. 9, when the difference between the first path 411 that the user moves according to the cutting exercise and the second path 421 determined by the cutting exercise is small, the first position 412 and the second Since the vector difference value of the two positions 422 is also small, the guide mark 404 may be displayed with a default size.

On the other hand, as shown in (b) of FIG. 9, when the difference between the first path 411 that the user moves according to the cutting exercise and the second path 421 determined by the cutting exercise is large, the first position 412 And since the vector difference value of the second position 422 is also large, the guide mark 404 may be displayed larger than the default size.

10 is a diagram for describing a display of a starting point, a first point, a second point, and a target point according to an exemplary embodiment.

Referring to FIG. 10, when the moving line moving from the starting point 401 to the target point 402 through the first point 501 and the second point 502 is set as cutting practice content, the starting point 401 ), the first point 501, the second point 502, and the target point 402 may be controlled to be displayed on the display unit 150.

First, when the device 100 detects that the user's hand is located at the starting point 401, the first point 501 is further displayed while the starting point 401 is highlighted, as shown in FIG. 10A. It can be controlled as much as possible.

When the device 100 detects that the user's hand position is out of the starting point 401, the first point 501 is highlighted and the second point 502 is further displayed as shown in FIG. 10B. It can be controlled to be displayed.

The device 100 controls the starting point 401 to no longer be displayed, as shown in (c) of FIG. 10, when the time out of the starting point 401 is determined to be greater than or equal to the reference time due to the movement of the user's hand position. can do.

The device 100 may predict the moving time to the first point 501 based on the first route 411 that is the user's moving route, and the user's hand to the first point 501 through the predicted moving time. Before this movement, the second point 502 can be controlled to be highlighted, as shown in FIG. 10D.

When it is detected that the user's hand is located at the first point 501, the device 100 may control the target point 402 to be further displayed, as shown in (e) of FIG. 10.

When the device 100 moves the position of the user's hand and the time out of the first point 501 is determined to be greater than or equal to the reference time, as shown in FIG. 10(f), the first point 501 is no longer It can be controlled not to be displayed.

The device 100 may predict the moving time to the second point 502 based on the first route 411 that is the user's moving route, and the user's hand to the second point 502 through the predicted moving time. Before this movement, the target point 402 can be controlled to be highlighted, as shown in FIG. 10G.

The embodiments described above may be implemented as a hardware component, a software component, and/or a combination of a hardware component and a software component. For example, the devices, methods, and components described in the embodiments are, 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 unit (PLU), microprocessor, or any other device capable of executing and responding to instructions. The processing device may execute an operating system (OS) and one or more software applications executed on the operating system. Further, the processing device may access, store, manipulate, process, and generate data in response to the execution of software. For the convenience of understanding, although it is sometimes described that one processing device is used, one of ordinary skill in the art, the processing device is a plurality of processing elements and/or a plurality of types of processing elements. It can be seen that it may include. For example, the processing device may include a plurality of processors or one processor and one controller. In addition, other processing configurations are possible, such as a parallel processor.

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 in a computer-readable medium. The computer-readable medium may include program instructions, data files, data structures, etc. alone or in combination. The program instructions 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. -A hardware device 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 not only machine language codes such as those produced by a compiler, but also high-level language codes that can be executed by a computer using an interpreter or the like. The hardware device described above may be configured to operate as one or more software modules to perform the operation of the embodiment, and vice versa.

The software may include a computer program, code, instructions, or a combination of one or more of these, configuring the processing unit to operate as desired or processed independently or collectively. You can command the device. Software and/or data may be interpreted by a processing device or, to provide instructions or data to a processing device, of any type of machine, component, physical device, virtual equipment, computer storage medium or device. , Or may be permanently or temporarily embodyed in a transmitted signal wave. The software may be distributed over networked computer systems and stored or executed in a distributed manner. Software and data may be stored on one or more computer-readable recording media.

As described above, although the embodiments have been described by the limited drawings, a person of ordinary skill in the art can apply various technical modifications and variations based on the above. For example, the described techniques are performed in a different order from the described method, and/or components such as systems, structures, devices, circuits, etc. described are combined or combined in a form different from the described method, or other components Alternatively, even if substituted or substituted by an equivalent, an appropriate result can be achieved.

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

Claims (3)

In a method for controlling quantitative cutting based on artificial intelligence, performed by a device,
Obtaining a 3D image of the group to be cut from the scanner unit when the group to be cut is scanned by the scanner unit;
Dividing the group to be cut into a plurality of areas based on the 3D image and calculating a volume for each area of the group to be cut;
When the weight of the to-be-cut entity is measured by a measuring unit, obtaining a weight of the to-be-cut entity from the measurement unit;
Calculating a weight for each area of the group to be cut based on the volume of the group to be cut and the weight of the group to be cut;
Outputting a control value for cutting the cut entity into a quantitative unit by applying the type, a unit of quantification, and the weight of each area to the artificial intelligence; And
And controlling the cutting unit to be cut in the quantitative unit by setting a blade type of the cutting unit, a cutting position, and a belt movement speed of the conveyor unit through the control value,
The step of calculating the volume for each area of the cut body,
Computing the volume of the group to be cut by summing the volume for each area of the group to be cut,
The step of calculating the weight for each area of the cut entity,
Calculating the density of the cut body by dividing the weight of the cut body by the volume of the cut body;
Determining which part of the group to be cut is a first area, which is one of the plurality of areas of the group to be cut, based on the 3D image;
If the first region is identified as a first region, obtaining a density of the first region from a database;
When the difference between the density of the cut body and the density of the first portion is identified within a predefined error range, the weight of the first region is calculated by multiplying the volume of the first region and the density of the cut body. step; And
If it is found that the difference between the density of the to-be-cut entity and the density of the first region is outside a predefined error range, the weight of the first region is obtained by multiplying the volume of the first region and the density of the first region. Comprising the step of calculating,
Quantitative cutting method based on artificial intelligence. delete The method of claim 1,
Obtaining a first image of a cut surface of the first cutting body, which is any one of the plurality of cutting bodies, when the cutting body is cut through the blade of the cutting unit and separated into a plurality of cutting bodies;
Classifying the first cut body into a plurality of zones according to the internal state of the first cut body based on the first image, and predicting and calculating a density for each zone of the first cut body;
Obtaining a price of the first state from a database when a first region, which is one of the plurality of regions of the first cutting body, is identified as a first state through density;
Calculating the weight in the first state by multiplying the weight of the first cut body and the ratio of the area occupied by the first area in the cut surface of the first cut body;
Calculating a price of the first zone by multiplying the weight of the first state and the price of the first state; And
The step of calculating the price of the first cutting body by summing the price calculated for each region of the first cutting body,
Quantitative cutting method based on artificial intelligence.

Images