KR102452080B1 - The system and method of determining rice grade and quality management using artificial intelligence - Google Patents

Abstract

The present invention relates to a system and a method of rice grade determination and quality management using artificial intelligence, which can quickly and accurately determine the rice grade using image processing by multiple CCD cameras and artificial intelligence. The system comprises a plate (2), an index table (4), a rice insertion hopper (5), a moisture meter (6), a plurality of rice supply units (8, 9), first and second cameras (11, 12), infrared cameras (13), first, second, and third ring lights (14, 15, 16), color chart units (21, 22, 23), a scraper (24), a roller brush (26), a brush motor (27), a dust collecting duct (28), a controller (29), and an artificial intelligence unit (30).

Description

The system and method of determining rice grade and quality management using artificial intelligence

The present invention relates to a rice grading and quality control system using artificial intelligence and a method therefor, and more particularly, by measuring the appearance quality of rice using a vision camera and artificial intelligence, and accurately grading, Comprehensive quality control of rice by measuring the whiteness, milling deviation, and ingredient data of rice, suggesting milling conditions to manage yield and taste, and comprehensively analyzing the appearance, quality, freshness (germination rate), ingredients, etc. was made to be possible.

In general, RPC (Rice Processing Complex) processes the entire process of rice individually, from import of harvested rice, through selection, weighing, quality inspection, drying, storage, and milling, to product shipment and sales, and by-product processing. Batch processing through a large-scale automation process rather than a farmhouse unit reduces labor force in rural areas, improves the quality of rice, improves the distribution structure, reduces costs, and supplements the purchasing function by reducing the amount of rice purchased. This is a significant contribution to market stability by absorbing a significant portion of the amount shipped during the harvest season in a situation where this is reduced by about 34%, brand rice is activated, and government purchases are decreasing.

On the other hand, the grain color sorting device is a foreign substance contained in grains (grains) such as rice, brown rice, glutinous rice, black rice, millet, barley, wheat, soybeans, etc. It is a rice processing complex device that distinguishes and separates grain, sand, synthetic resin or glass. After classifying into (including defective products), foreign substances are separated and removed by instantaneous jetting of high-pressure air to separate and treat good products and defective products.

On the other hand, high-quality rice can be defined as rice with good grain, no foreign matter, excellent transparency and gloss by completely removing the appropriate moisture content and husk layer, and rice with a lot of intrinsic flavor or rice with appropriate ingredient standards. The main factors that determine the quality of rice are appearance and ingredients, which can be divided into transparency, gloss (gloss), whiteness, moisture content, ingredients (protein, amylose, etc.), and appearance (sizing, sizing, grain, etc.). These factors has a very high correlation with the taste of rice, so the government authorities set standards for the quality of rice, and the quality grade determined according to the standard is indicated on the rice container or packaging, but there are many errors in the grade determination. In the process of shape recognition and grading test, many rice shapes with ambiguous grading were found, which is a reality that requires technical supplementation.

Republic of Korea Patent Publication No. 10-1189600 (Title of the invention: Rice quality protein automatic determination machine, 2012. 10. 10. Patent announcement) Republic of Korea Patent Publication No. 10-2001-0104795 (Title of the invention: Crude oil grade determination system, 2001. 11. 28. Patent publication) Republic of Korea Patent Publication No. 10-1808461 (Title of the invention: Method and apparatus for predicting the remaining life of a machine, 2017. 12. 12. Patent announcement)

The problem to be solved by the present invention is to determine the grade by measuring the appearance quality of rice using a Vision CCD Camera and artificial intelligence, and to measure the whiteness and grinding deviation of the rice to present the milling conditions to determine the yield and taste A rice grading and quality management system and method using artificial intelligence that enables comprehensive quality management of rice by providing a taste value by comprehensively analyzing the appearance, quality, freshness (germination rate), and ingredients of the rice. is in providing.

The rice grading and quality control system using artificial intelligence of the present invention includes a plate installed on the top of the base, a transparent index table axially installed in the center of the plate and rotated in one direction by a servo motor, and a rice hopper installed on the base A moisture meter installed in the rice input hopper and input to the artificial intelligence unit by measuring the moisture of the input rice, a plurality of rice supply units installed at the bottom of the rice input hopper and supplying the rice to the index table upper surface, and installed on the base The first camera, which is installed on the base, is installed on the base, and captures the transmitted light from the rice, and measures the appearance, quality, whiteness, sage, perfect grain, harm grain, and milling deviation of the rice by imaging the reflected light from the rice. A second camera that measures and inputs into the artificial intelligence unit, installed on the base, and captures the reflected light from the rice to measure the protein, amylose, fat, carbohydrate, etc. of the rice, and then input infrared rays (including near infrared rays) to the artificial intelligence unit. ) a camera, a first ring light installed between the first camera and the index table to illuminate the rice, a second ring light installed below the index table where the second camera is positioned to illuminate the rice, an infrared camera and an index The third ring lighting lamp installed between the tables and illuminating the rice with infrared light, rotates at right angles by each step motor and moves to the lower part of the first and second cameras and the infrared camera to provide standard illumination for each camera and then rotates back. The color chart unit, the scraper installed on the base and moving the rice to the discharge hopper located at the edge of the index table, and the roller brush installed on the base and adsorbing and cleaning rice bran, fine powder, and foreign substances on the upper surface of the index table and a brush motor, a dust collecting duct installed on the upper part of the roller brush and collecting rice bran, fine powder, and foreign substances adsorbed on the roller brush, and an artificial intelligence unit connected to the controller and the controller.

The rice input part is disposed at the 0 ° position of the index table, and counterclockwise based on this, the first camera is at a position of 60 °, the second camera is at a position of 120 °, an infrared camera is located at a position of 180 °, and a scraper is located at a position of 240 °, It may be characterized in that a roller brush is disposed at the 300° position.

The artificial intelligence unit includes a data collection unit for collecting rice moisture data input through a moisture meter and rice image data input from the first and second cameras and infrared cameras, a database for storing the collected data, and processing the collected data It includes a data processing unit, an artificial intelligence learning unit that learns from the processed data, and a determination unit that determines a rice grade based on the learned data, and presents the whiteness, grain deviation, and component data of the rice.

The index table is preferably made of transparent tempered glass having a predetermined thickness and a predetermined diameter to enable transmission illumination.

The first and second cameras and the infrared camera may be a charge-coupled device camera (CCD camera).

The first and second cameras and the infrared camera may include an image processing module and an automatic color & illuminance correction algorithm.

It further includes a horizontal bar for evenly spreading the rice supplied to the index table on the upper surface of the index table and maintaining it as one layer.

The artificial intelligence algorithm of the artificial intelligence unit may be any one of Deep Learning, CNN, ClassDistributions, DecisionTree, GradientBoostedTrees, LogisticRegression, Markov, NaiveBayes, NearestNeighbors, NeuralNetwork, RandomForest, and SupportVectorMachine.

The first, second, and third ring lights are provided with a ring-shaped PCB on which a plurality of LEDs are mounted at predetermined intervals on the inner inclined surface of a circular ring having a through hole in the center, and the illumination light is focused toward a point on the central virtual line of the through hole. It may be characterized in that it has a predetermined slope to be able to be.

The first and third ring lights are installed on the index table so that the reflected light of the rice is incident on the first camera and the infrared camera, respectively, and the second ring lights are installed on the lower part of the index table so that the transmitted light of the rice is incident on the second camera can be characterized.

The present invention provides a rice grade determination and quality control method using artificial intelligence, comprising: a) rotating the color chart unit at right angles by a stepping motor to move to the lower portions of the first, second and infrared cameras, respectively; b) first and second A camera and an infrared camera image the surface of the color chart to correct the reference illuminance, c) to return the color chart to the outside of the index table by a stepping motor, d) to feed the rice into the rice hopper. Measuring moisture with a moisture meter and then inputting it into the artificial intelligence unit, e) supplying the rice from the rice hopper to the upper surface of the index table rotating in one direction through the rice input unit, f) the lower part with the first camera Measuring the appearance, quality, whiteness, sage, perfect grain, harmless grain, milling deviation, etc. by imaging the reflected light of the rice moved to The steps of measuring the granular grain by imaging the transmitted light and inputting it to the artificial intelligence unit, h) capturing the infrared reflected light reflected from the rice moved to the lower part of the infrared camera Measurement and input to the artificial intelligence unit; i) dropping and discharging the rice moved by the scraper into the discharge hopper at the edge of the index table; j) Residual flour, rice bran, foreign matter, etc. The step of collecting dust with the dust collecting duct on the upper part of the roller brush after adsorbing it; and judging the grade of the rice.

In step e) in which the rice of the rice input hopper is supplied to the upper surface of the index table rotating in one direction through the rice input unit, the rice supplied to the index table is evenly spread by the horizontal bar and moving along the index table further includes

The artificial intelligence unit learns the data input by the moisture meter and the first, second, and third cameras with the rice grade learning unit, and in step k), the artificial intelligence unit learns the whiteness of the rice, It further comprises the step of presenting the calibration deviation and component data (step l).

After the step k) in which the artificial intelligence unit learns the data input by the moisture meter and the first, second, and third cameras with the rice grade learning unit, and determines the grade of the rice as special, prize, normal, etc., the artificial intelligence unit It further includes the step of presenting the whiteness, the polishing deviation, and the component data together.

The artificial intelligence unit learns the data input by the moisture meter and the first, second, and third cameras with the rice grade learning unit, weighs and packs it after step k) to determine the grade of rice as special, prize, normal, etc. and marking the grade on the packaging surface.

The present invention measures moisture, ragi, damaged grains, heat-damaged grains, other foreign matter, and powdery grains (6 items, see Table 1) for grading rice (polished rice) with images, and grades rice according to the government quality determination standard It can be accurately judged as special, superior, or normal.

The present invention can determine the quality of brown rice by measuring moisture, scallops, damaged ribs, damaged ribs, and other foreign substances with an image for judging the quality of brown rice.

The present invention can measure the whiteness and milling deviation (CBB) of the rice to determine whether it exceeds or falls short of 45 degrees, which is the proper milling standard for rice, and can measure the milling deviation of the processed rice, that is, the degree of uniform milling.

In the present invention, the freshness of rice can be determined by predicting the germination rate by image and component measurement by a CCD camera, and the component measurement of rice can be measured using spectroscopic analysis (NIR) using an infrared camera.

The present invention uses image processing by a plurality of CCD cameras and an artificial intelligence algorithm to measure the appearance quality of rice to determine the grade, and to measure the whiteness and polishing deviation of the rice to present the milling conditions to determine the yield and It manages the taste and provides the taste value by comprehensively analyzing the appearance quality, freshness (germination rate), and ingredients of the rice, thereby achieving comprehensive quality control of rice.

Rice grading and quality management system using artificial intelligence of the present invention uses image processing and artificial intelligence by the first and second cameras 11 and 12 and infrared cameras 13 moving along the index table 4 This has the effect of quickly and accurately determining the grade of the rice (7).

In the present invention, a plurality of LEDs 43 constituting the first, second, and third ring lighting lamps 14, 15, and 16 are installed slowly (a point O on the central virtual line of the through hole 41) It is installed so that it can be focused toward) and the rice 7 on the index table 4 is intensively illuminated, so that the illumination light of the LED 43 is directly incident on the first and second cameras 11 and 12 and the infrared camera 13. Not only does it not occur, but also factors that interfere with image processing due to the shadow effect are removed, so there is an effect that can be measured without image loss such as color or shape of the surface of the rice grain 7 .

The present invention uses the infrared and/or near-infrared light from the third ring light 16 and the infrared camera 13 to provide protein, amylose, fat, carbohydrate, etc. has a measurable effect.

The present invention detects the difference in internal density, internal cracks, and transmittance as well as the granularity of the rice 7 by backlighting by the second ring lamp 15 installed under the index table 4, and through this, the rice ( 7) has the effect of detecting the condition.

In the present invention, the illumination correction of the first and second cameras 11 and 12 and the infrared camera 13 is periodically performed or The accuracy is maximized by setting the first and second cameras 11 and 12 and the infrared camera 13 according to the ambient illuminance atmosphere by performing the measurement at the beginning of the measurement or when the rod of the rice 7 is changed. has the effect of being

The present invention has the effect of determining the cause of the milling deviation by measuring external characteristics such as the long and short axis of the grain using an image processor.

The present invention has the effect of directly displaying the determined grade on the surface of the rice wrapping paper by interlocking with an inkjet printer, a marker, etc. with accurate grade determination of rice.

According to the present invention grade determination result, there is an effect that it can be stored and managed in tanks classified as “special, high, and normal”, and according to the mixing ratio of the saffron, the sarragi sorter (vibrating screen, rotary shifter, whether to operate, automatic control interface) By selectively controlling the

The present invention is a rice cleaning machine by utilizing the obtained data such as quality, whiteness, and polishing deviation. It is possible to control and manage (automatic feedback control/artificial intelligence cleaning system) of the rice-cleaning machine under the optimal grinding conditions, such as manual or automatic control of the rice-polishing machine, pressure control of the rice-polishing machine and polishing room, motor current value control, and motor RPM control. This has the effect of greatly contributing to the improvement of the yield and quality of rice.

The present invention improves the production yield, production cost, and quality of rice produced in grain processing plants such as national rice processing plants (RPC), agricultural cooperative corporations, and rice processing companies so that rice processing and distributors can produce high-quality rice And it is a very useful invention that has effects such as maximizing profits.

Figure 1: A perspective view of a rice grade determination and quality control system shown as an example of the present invention.
Figure 2: A plan view of the rice grading and quality control system shown as an example of the present invention.
Figure 3: A front view of the rice grading and quality control system shown as an example of the present invention.
Figure 4: A side view of the rice grading and quality control system shown as an example of the present invention.
Figure 5: a plan view (or bottom view) of the ring light shown as an example of the present invention.
Figure 6: A cross-sectional view of a ring light shown as an example of the present invention.
7 is a plan view of the use state of the rice grading and quality management system shown as an example of the present invention.
8 : A plan view of the use state of the rice grading and quality management system shown as an example of the present invention, with the first and second cameras and the infrared camera removed.
Figure 9: A circuit configuration diagram of the rice grading and quality management system shown as an example of the present invention.
10 : A block diagram of an artificial intelligence unit shown as an example of the present invention.
Figure 11: A process flow chart of the rice grading and quality control system shown as an example of the present invention.
12 : Photographs of various shapes of rice.

Hereinafter, preferred embodiments of the present invention will be described in detail with reference to the accompanying drawings. In describing the embodiments of the present invention, the same components in the drawings are denoted by the same reference numerals as much as possible, and detailed descriptions of related known configurations or functions are omitted so as not to obscure the gist of the present invention, and, in addition, the accompanying drawings Matters expressed in may be different from those actually implemented in the drawings schematically for easy explanation of the embodiments of the present invention.

1 is a perspective view of a rice grading and quality control system using artificial intelligence shown as an example of the present invention, FIG. 2 is a plan view thereof, and FIG. 3 is a front view thereof, which is installed on top of a base 1 having a predetermined height Plate (2), an index table (4) axially installed in the center of the plate (2) and rotated in one direction by a servo motor (3), a rice input hopper (5) installed on the upper part of the base (1), and rice A moisture meter (6) installed in the input hopper (5) and measuring the moisture of the input rice, and a plurality installed under the rice input hopper (5) and discharging the rice (7) to the upper surface of the index table (4) It is installed on the rice supply part 8 and 9 and the base 1 of the The first camera 11, which measures and then inputs to the artificial intelligence unit 30, and the second camera ( 12) and an infrared (including near-infrared) camera that measures the protein, amylose, fat, carbohydrate, etc. of the rice 7 by imaging the reflected light of the rice 7 and then inputs it to the artificial intelligence unit 30 ( 13), the first ring lighting lamp 14 installed between the first camera 11 and the index table 4 to illuminate the rice 7, and the index table 4 on which the second camera 12 is located A second ring lighting lamp 15 installed under the , 2 cameras 11 (12) and infrared cameras 13 are installed below and rotated at right angles (90°) by respective step motors 17, 18 and 19 to each camera 11 and 12 The color chart parts 21, 22, and 23 that return (0°) rotate after providing the reference illuminance with (13), and the rice installed on the base (1) and judged on the edge of the index table (4). A scraper 24 for moving and discharging to the located discharge hopper 25, and an index table 4 installed on the base 1 ) A roller brush 26 that absorbs and cleans rice bran, fine powder, and foreign substances on the upper surface, a brush motor 27 that rotates the roller brush 26 at high speed in one direction, and a roller installed on the top of the roller brush 26 A dust collecting duct 28 and a controller 29 for dust collecting treatment of rice bran, fine powder, foreign substances, etc. adsorbed to the brush 26, and an artificial intelligence unit 30 connected to the controller 29 and having a rice determination learning unit are further added. include

It further includes a horizontal bar 10 that evenly spreads the rice (rice) 7 supplied to the index table 4 on the upper surface of the index table 4 and maintains it as one layer.

The rice input hopper 5 may have an appropriate amount, for example, 100 to 2,000 g of rice 7 at a time, and a moisture meter 6 is installed in the rice input hopper 5 to input rice. After measuring the moisture in (7), it is input to the artificial intelligence unit (30).

The first and second cameras 11 and 12 and the infrared camera 13 are charge-coupled device (CCD) cameras, and include an image processing module and an automatic color & illuminance correction algorithm.

A color sheet is installed on the upper surface of the color chart units 21, 22, and 23 in a detachable structure to provide reference illuminance to each CCD camera 10, 11, and 12.

The moisture meter 6 may be a non-destructive moisture meter, such as an electrical resistance moisture meter, a laser moisture meter, a near-infrared moisture meter, and the like.

As the artificial intelligence algorithm in the present invention, a deep learning algorithm, CNN, ClassDistributions, DecisionTree, GradientBoostedTrees, LogisticRegression, Markov, NaiveBayes, NearestNeighbors, NeuralNetwork, RandomForest, SupportVectorMachine, etc. may be mentioned as examples.

The first, second, and third ring lights 14, 15, and 16 have a hole in the center so that the rice grain 7 on the index table 4 can be imaged with a camera positioned on the upper part as shown in FIGS. 5 and 6 . It has a circular ring 44 shape in which 41 is formed, and a ring-shaped PCB 42 on which a plurality of LEDs (Light Emitting Diodes) 43 are mounted at predetermined intervals is installed on the inner inclined surface of the circular ring 44. , has a predetermined inclination so that the illumination light can be focused and illuminated toward the central virtual line point O of the through hole 41 where the rice grain 7 is located.

Therefore, not only the illumination light is not directly incident to the first and second cameras 11 and 12 and the infrared camera 13, but also factors that interfere with image processing due to shadow effects are removed, so the color or shape of the surface of the rice grain 7, etc. can be measured without image loss of

The first and third ring lighting lamps 14 and 16 are installed at a predetermined height above the index table 4 so that the reflected light of the rice 7 is incident on the first camera 11 and the infrared camera 13, respectively, and is photographed. , The second ring lighting lamp 15 is installed under the index table 4 and illuminates to pass through the rice 7 at the top so that the transmitted light of the rice 7 is incident to the infrared camera 13 .

That is, since the rice 7 is intensively illuminated by the first, second, and third ring lighting lamps 14, 15, and 16, the surface reflected light of the rice 7 passes through the through hole 41 and the first and third cameras 11 ) (13) is incident on the image, and the transmitted light passing through the rice grain (7) is incident on the second camera (12) and imaged.

The roller brush 26 has a configuration in which thin wool or filament yarns are tightly fixed to a predetermined length on the outer circumferential surface of the rod body, and the index table 4 is rotated at high speed in the opposite direction to which the index table 4 is moved by the brush motor 27 while rotating at high speed. ) The fine powder, rice bran, foreign material, etc. remaining on the surface are swept and adsorbed, and the adsorbed dust and small fine powder are removed by vacuum suction through the suction port of the dust collecting duct (28).

In the rice grade determination and quality control system using artificial intelligence of the present invention, the transparent index table 4 rotates at constant speed in one direction by the servo motor 3 while the first and second cameras 11 and 12 and the infrared camera 13 ) and artificial intelligence to measure the appearance quality of rice to make accurate grade determination, measure the whiteness, milling deviation, and ingredient data of the rice to suggest milling conditions to manage yield and taste, and to Germination rate), ingredients, etc. are comprehensively analyzed to provide sikmichi, so that comprehensive quality control of rice is made.

The present invention is another embodiment, the rice input unit (8) (9) and the first camera (11) at intervals of 60 ° around the index table (4) rotated step (Step) in one direction by the servo motor (3) Wow, the second camera 12, the infrared camera 13, the scraper 24, and the roller brush 26 are sequentially installed so that each step process or operation can be collectively performed.

The index table 4 rotates at an angle of 60° (6 steps or divided by 60°) by the servomotor 3 controlled by the controller 29 , and grades and discharges the rice 7 supplied.

For example, the rice inserting units 8 and 9 are disposed at the 0° position of the index table 4, and the first camera 11 is positioned at 60° counterclockwise based on this, and the second camera is located at the 120° position. (12), the infrared camera 13 at the 180° position, the scraper 24 at the 240° position, and the roller brush 26 at the 300° position are installed.

Here, the rice input units 8 and 9 supply the rice 7 fed into the upper rice input hopper 5 to the upper surface of the index table 4 in at least one row or more, preferably in two rows, and the first The camera 11 captures the reflected light reflected from the rice 7 using the first ring lamp 14 illuminating the top of the rice 7, but the appearance, elegance, whiteness, saragi, and full grain of the rice 7 , damage grain, sizing deviation, etc. are measured and input to the artificial intelligence unit 30, and the second camera 12 uses the second ring lighting lamp 15 positioned below the rice 7 to display the rice 7 The transmitted light is captured, but the granular grain is measured and input to the artificial intelligence unit 30, and the infrared camera 13 uses the third ring lighting lamp 16 that illuminates the upper portion of the rice 7 with infrared light. After measuring the internal components of (7), such as protein, amylose, fat, carbohydrate, etc., it is input to the artificial intelligence unit 30, and the scraper 24 indexes the rice (7) that is moved after the grading is finished. It moves and falls to the discharge hopper 25 located at the edge of the table 4 to discharge it to the outside of the index table 4, and the roller brush 26 is a It plays the role of cleaning by adsorbing rice bran, fine powder, and foreign substances.

Accordingly, the above processes are sequentially performed while moving to the next step by rotating the index table 4 by 60° for each step by rotation of the index table 4, and, of course, the sequential processes are performed at once.

On the other hand, in the present invention, the stepping motors 17, 18, and 19 installed around the index table 4 rotate at right angles to the first, 2 The cameras 11, 12 and the infrared camera 13 are moved to the lower part, respectively, and the first and second cameras 11 and 12 and the infrared camera 13 are displayed in the color chart unit 21, 22, and 23. After imaging the surface of , the illuminance is corrected using the automatic color & illuminance correction algorithm, and the color chart units 21, 22, and 23 are After returning to its original position outside the index table 4, a predetermined amount of rice 7 is put into the input hopper 5, and the current moisture of the rice in the input hopper 5 is measured by the moisture meter 6 At least one row is input to the artificial intelligence unit 30, and the input rice 7 is supplied to the surface of the index table 4 that rotates stepwise in one direction through the rice input unit 8 and 9, and the index table 4 ) moved along the index table 4 with the rice supplied to the horizontal bar 6 spread evenly by the horizontal bar 6, and the reflected light of the rice 7 was captured by the first camera 11 to capture the appearance of the rice 7, After measuring the quality, whiteness, sage, perfect grain, damage grain, and grain deviation, etc., it is input to the artificial intelligence unit 30, and the transmitted light of the rice 7 is captured by the second camera 12 to measure the grain quality. Then, it is input to the artificial intelligence unit 30, and the infrared camera 13 captures the reflected light of the rice 7 to measure the protein, amylose, fat, carbohydrate, etc. of the rice 7, and then to the artificial intelligence unit ( 30), and the rice 7, which is moved after grading is finished, is moved to the discharge hopper 25 located at the edge of the index table 4, and it is responsible for discharging it to the outside of the index table 4 by dropping it. The rotating roller brush 26 sweeps and adsorbs rice bran, fine powder and foreign substances remaining on the upper surface of the index table 4 to clean it, and the dust collecting duct 28 above the roller brush 26 collects fine powder, rice bran, and foreign substances adsorbed on the roller brush (26). Since it is processed, the upper surface of the index table 4 is always cleanly cleaned, so that it is possible to continuously determine the rice grade.

9 is a circuit configuration diagram of a rice grading and quality control system shown as an example of the present invention, a moisture measuring system 6 measuring and inputting the moisture of the rice to the input of the controller 29, the reflected light of the rice 7 and The first and second cameras 11 and 12 and the infrared camera 13 that capture transmitted light image processing and then input are connected, and a setting unit S for selecting or setting an operation state, etc. is connected, and the controller 29 In the output of the display unit (D) for displaying various operation and error states and data, the first, second, and third ring lights (14, 15) (16) to illuminate the rice (6), and the index table (4) A servo motor 3 that rotates the index, step motors 17, 18, 19 that drive the color charts 17, 18, 19, and a brush that rotates the roller brush 26 It includes a motor 27, a memory unit M, an artificial intelligence unit 30 connected through an interface, and a rice processing plant (RPC) connected to the artificial intelligence unit 30 .

It may be connected to a grain processing plant such as an agricultural cooperative corporation or a rice processing company, as well as a rice processing plant (RPC) connected to the artificial intelligence unit 30 .

The controller 29 may be an artificial intelligence control unit including the artificial intelligence unit 30 .

10 is a block diagram of the artificial intelligence unit 30 shown as an example of the present invention, and the moisture data of the rice 6 input through the moisture meter 6 and the first and second cameras 11 and 12 and the infrared camera ( 13) a data collection unit 31 for collecting image data of rice 6 input from the input, a database 35 for storing and reading the collected data, and a data processing unit 32 for processing the collected data; , an artificial intelligence learning unit 33 that learns from the processed data, and a determination unit 34 that determines a rice grade based on the learned data, and presents the whiteness, milling deviation, and component data of the rice 7 do.

11 is a flow chart of rice grade determination shown as an example of the present invention, in which the color chart units 21, 22, and 23 are rotated at right angles by the stepping motors 17, 18, and 19 to rotate the first and second cameras ( 11) (12) and the infrared camera 13 move to the lower part (step a), respectively, and the first and second cameras 11 and 12 and the infrared camera 13 are moved to the color chart unit 21, 22 ( 23) is imaged to set a reference color necessary for grading (step b), and the color chart units 21, 22, and 23 are indexed by the stepping motors 17, 18, and 19. (4) It returns to the outside position (step c), and the moisture data measured by the moisture meter 6 of the rice 7 fed into the rice hopper 5 is input to the artificial intelligence unit 30 ( Step d), the rice 7 of the rice input hopper 5 is fed to the upper surface of the index table 4 that rotates in one direction through the rice input parts 8 and 9 (step e), and the first camera (11) Take a picture of the reflected light of the rice (7) moved to the lower part, measure the appearance, elegance, whiteness, sage, perfect grain, harmless grain, milling deviation, etc., and input it into the artificial intelligence unit 30 (step f) , the second camera 12, by imaging the transmitted light of the rice grain 7 moved to the lower part to measure the granularity, and then input it to the artificial intelligence unit 30 (step g), and the infrared camera 13 moved to the lower part By imaging the infrared reflected light of the rice 7, the protein, amylose, fat, carbohydrate, etc. of the rice 7 are measured, then input to the artificial intelligence unit 30 (step h), and moved to the scraper 24 The rice 7 is dropped and discharged into the discharge hopper 25 at the edge of the index table 4 (step i), and the fine powder, rice bran, foreign substances, etc. remaining on the surface of the index table 4 moved to the roller brush 26 are adsorbed. Then, the dust is collected by the dust collecting duct 28 of the upper part of the roller brush 26 (step j), and the data input to the moisture meter 6 and the first, second, and third cameras 11, 12, and 13 are artificially collected. The intelligent unit 30 learns the rice grade learning unit and determines the grade of the rice as special, prize, normal, etc. (step k).

The artificial intelligence unit 30 learns the data input from the moisture meter 6 and the first, second, and third cameras 11, 12, and 13, and determines the grade of rice as special, prize, normal, etc. After the step k), the artificial intelligence unit 30 further includes the step of presenting the whiteness, grain deviation, and component data of the rice.

In step e), the rice 7 of the rice input hopper 5 is supplied to the upper surface of the index table 4 rotating in one direction through the rice input unit 8 and 9, the index table 4 It further includes the step of moving along the index table (4) in a state in which the rice supplied to the horizontal bar (10) is spread evenly.

The artificial intelligence unit 30 learns the data input from the moisture meter 6 and the first, second, and third cameras 11, 12, and 13 as a rice grade learning unit, so that special, prize, normal, and other types of rice are classified. After the step k) of determining the grade of , the method further includes the step of measuring the packaging and then displaying the determined grade on the packaging surface.

The camera illumination correction, rice grade determination, rice germination rate measurement, rice rice evaluation, rice milling deviation measurement, rice quality measurement, etc. of the present invention will be described below.

[Camera Illumination Correction]

Illuminance correction using the color chart units 21, 22, and 23 of the present invention can be performed periodically or at the beginning of measurement, or when the rod of the rice 7 is changed, etc., the first and second cameras (11) (12) and infrared camera (13) are adjusted to the ambient illuminance atmosphere, the rice measurement error is significantly reduced.

[Rice grade judgment]

The present invention relates to the moisture (moisture content) of the rice 7 measured by the moisture meter 6, and the image obtained by processing the image of the rice 7 captured with the first and second cameras 11 and 12. The grade of the rice 7 is judged as "special", "superior", "normal", and "extra" by the ratio of damaged grain, other foreign matter, and powdery grain (see Table 1).

The rice grade standards according to the enforcement regulations of the Korean Grain Management Act are exemplified in Table 1 below.

division maximum limit moisture fight granular granulation avoid heat loss other foreign matter special
16% 3% 2% One% 0% 0.1% award 7% 6% 2% 0% 0.3% usually 20% 10% 4% 0.1% 0.6%

For other conditions, heat loss should be less than 3 grains in 1 kg of sample, 'special' should be less than 7 grains, and among other foreign substances, 1 kg of sample should be 3 The number of repeated counts and summation should be less than 1, and in the case of different grains (including nui), 'Special' and 'Award' are 2 or less, and 'Normal' is 5 or less. ', 'Normal' can be marked (marked), and it is stipulated that 'Head Rice' can be marked on the cover sheet separately from the 'Special' mark only when the complete grain ratio is 96.0% or more. 12 is a photograph showing the types and shapes of grains, such as full grains, heat loss grains, damage grains, granular grains, saragi, and donghwal, and is defined as follows.

- Percentage (%): It refers to the weight ratio to the whole quantity, and is rounded to two decimal places.

- Moisture: It refers to the moisture content measured by the 105℃ drying method or a method that can obtain an equivalent result.

- Saragi: Among the grains that do not pass through the sieve, the nominal dimension of 1.7㎜ among KS A 5101-1 (Metal Mesh sieve) refers to those whose length is less than 3/4 of the average length of the complete grain.

- Powdery granularity: It refers to grains with more than 1/2 of the volume in a powdery state.

- Damaged grains: Contaminated grains, diseased grains, insect pests, germinated grains, menstrual intestinal free grains, red algae and black algae are attached to more than 1/4 of the length of the grain. However, minor damage that does not affect the quality of rice is excluded.

- Heat-damaged grain: Refers to grains that are discolored or damaged by heat, etc., and more than 1/4 of the surface area of fine grains is colored orange (more than 2.5Y 8/4 standard of Korean standard color sampling). However, if the degree of coloration is less than the orange standard or less than 1/4, it is applied as a non-damaged grain.

- Other foreign substances: Refers to those products other than rice (solid substances such as stones, plastics, glass, iron chips, different grains) and KS A 5101-1 (metal meshes) that have been passed through a sieve by hitting them with a metal mesh with a nominal size of 1.7 mm.

- Heterogeneous grain: refers to grains other than rice (including nui).

- Perfect grain: Refers to grains that are perfect grains or 3/4 or more of the average length of complete grains due to the appearance of rice, excluding powdery grains, damaged grains, and damaged grains. The average length of grains is 15 or more complete grains. Measure and calculate.

[Measurement of rice germination rate]

In rice, gibberellin, a plant growth material, is secreted under moderate temperature and moisture (30-35%), and biosynthesis occurs in the gene of the blastocyst epithelial tissue, resulting in massive replication of hydrolytic enzymes and carbohydrate transfer from the endosperm to the embryo. As they move in large amounts, respiration and metabolism become active, resulting in increased dry weight and germination.

The ratio of seeds that can germinate among a certain amount of rice is called the germination rate, and the method of measuring the germination rate of rice is to measure the number of germinated 100 grains within 7 days at a temperature of 20. In addition, in order for germination to occur, the enzyme must be activated in the embryo, and the absorbance can be measured in a specific wavelength band by applying spectroscopy to the enzyme activity. That is, although there is a difference in the degree of general raw material conditions, degradation of quality by enzymes continues to occur in the same way as in the germination process, and atomic group and trace component content can be measured by irradiating light in various wavelength bands.

Therefore, the high germination rate in rice means that the ratio of live embryos is high, and the fact that the rice is alive means that it has enzyme activity. Therefore, measurement of enzyme activity, i.e., the measurement of metabolites generated by enzyme reaction, can be measured by spectroscopic analysis. do.

In the present invention, it is possible to measure the germination rate of rice (7) with improved accuracy by using image processing and artificial intelligence algorithms by the infrared camera (13). The germination rate of the rice 7 can be more accurately measured by applying the spectroscopic analysis data of the rice 7 and the shape characteristic and color characteristic data of the embryo part to an artificial intelligence algorithm.

In the present invention, the germination rate of rice 7 is obtained by capturing an image of rice 7 with an infrared camera 13, and then extracting the contour of the embryo area in consideration of color, position and shape characteristics to recognize the shape The germination rate can be measured by analyzing the area, circumference, short axis, long axis, long and short axis ratio, outline ratio, circularity, and color factor (Lab) of the embryo.

For example, the infrared and/or near-infrared reflected light reflected from the rice 7 by the infrared and/or near-infrared illumination light irradiated from the third ring lighting lamp 16 to the rice 7 is sent to the infrared camera 13. After detecting the image by imaging, the shape characteristic and color characteristic data of the rice 7 obtained through image analysis may be applied to the artificial intelligence algorithm of the artificial intelligence unit 30 to measure the germination rate of the rice 7 .

The third ring lighting lamp 16 is installed on the transparent turntable 4, and is placed on the upper surface of the transparent turntable 4 to step along the transparent turntable 4 to the lower portion of the rice grain 7, infrared and/or near-infrared rays. will illuminate

The infrared camera 13 is a CCD camera capable of recognizing the spectrum of the near-infrared (NIR) band, detects the ultraviolet and/or near-infrared light spectrum from the reflected light from the rice 7, and photoelectrically converts it to the rice 7 get an image

The shape characteristics obtained from the image of the rice (7) are the area (area), perimeter (perimeter), minor axis, major axis (major axis), major and minor axis ratio, contour ratio of the rice (7) embryo part. ) and at least one selected from the group consisting of roundness, and the color characteristic may be a color factor (Lab) value of the embryo part of the rice (7).

In the present invention, in order to measure the germination rate of the rice 7, the rice 7 is illuminated with a light source in the ultraviolet (UV), visible and near infrared (NIR) regions with the third ring lighting lamp 16, and the rice 7 ), an ultraviolet light spectrum of a 100-500 nm band and a near-infrared light spectrum of a 700-2,000 nm band may be detected by the infrared camera 13 in the reflected light.

The image obtained from the infrared camera 13 can measure the germination rate of the rice 7 by applying the shape characteristics and color characteristics data of the rice 7 obtained by spectroscopic analysis and image analysis to the germination rate prediction program based on the artificial intelligence algorithm. have.

[Rice taste evaluation]

On the other hand, the quality and taste of rice vary greatly depending on various conditions in the post-harvest process such as import, drying, storage and milling as well as the pre-harvest variety and cultivation conditions.

That is, water, protein, lipid, carbohydrate, crude fiber, and crude flour are the main components of rice, and the ratio of the components varies depending on the drying, storage and milling method as well as the variety or cultivation environment. Carbohydrates, which account for about 75% of the total weight of rice, are starch composed of amylose and amylopectin. It is classified into high long-grain species (30-40%) and the like. It is known that the amylose content is closely related to the viscosity of rice, so the higher the content, the lower the viscosity and the higher the hardness, the lower the taste.

Protein is the second most abundant component of rice after starch, and contains about 74% of brown rice and 68% of white rice. It is known that the taste is low.

White rice contains about 1% of lipids composed of neutral lipids whose main fatty acids are oleic acid and linolenic acid. During storage of white rice, fat combines with oxygen in the air to easily hydrolyze or oxidize, resulting in an increase in aldehydes and ketones such as hexanal and pentanal, creating a bitter taste, and the stickiness and taste of rice will also have a bad effect on Hydrolysis acts on the ester bond of lipids to produce free fatty acids, and it is known that various enzymes such as lipase and lipoxidase are involved, especially when the temperature and moisture content are high. goes fast In general, the fatty acid value, which is an indicator of fatty acid oxidation, is used as one of the important criteria for judging whether grains are safely stored. In addition, a variety of factors such as moisture content, Mg/K ratio and degree of hardness, and appearance characteristics affect taste.

Factors influencing the various taste factors of rice include varieties, production areas according to weather and soil, production year, cultivation methods such as nitrogen fertilization, harvest time, etc. ) Post-harvest (post-harvest) that can determine the taste of white rice, which is the final product, such as processing method, drying temperature, drying time, etc., storage method such as storage temperature and storage period, and polishing method such as polishing degree and embryo adhesion rate ) can be classified by the treatment method.

The components determined by the pre-harvest treatment method, namely, amylose, protein and lipid, have different ratios depending on the degree of grinding or dry storage method. . In addition, depending on the drying and storage method, the lipid changes and the oxidation of fat, that is, the fatty acid value, increases significantly due to hydrolysis. After about 14 weeks, it approaches 15 mLL.

The method of evaluating the taste of rice is largely a sensory evaluation by the sense that a person evaluates the taste, aroma, appearance, texture, etc. by tasting the rice, and by measuring the taste influencing factors by physicochemical or mechanical methods, it is evaluated by a specific taste. can be distinguished in a way that The sensory taste evaluation is a direct method to comprehensively evaluate the taste of rice, but it requires a lot of manpower, time, and samples for evaluation, and different results can be derived depending on individual preferences. In recent years, taste factors affecting rice have been identified, and as various physicochemical/mechanical methods to measure these factors have been developed, a food taste evaluation method using them is being introduced. In order to evaluate taste by such a physicochemical/mechanical method, the precision of the food formula using the food taste factor that can express the taste as a variable is a very important indicator.

The formula with the highest crystallinity among the food formulas published so far is the food formula described in the paper published by Horino et al. However, this formula is a formula developed with minerals as a variable, and it is difficult to apply it in production sites such as RPC or distributors because it requires a long time-consuming physicochemical method to measure minerals.

As described above, the taste of the rice 7 is determined according to the pre-harvest treatment method such as the variety, production area and cultivation method, as well as the post-harvest treatment method such as drying, storage and processing, etc. A food formula and evaluation method taking into account

[Measurement of grain milling deviation]

The dictionary meaning of milling is 'removing the husk and rice bran layer of rice', and it has the same meaning as processing (加工, processing), 'to make new things by putting more hands on raw materials or materials'. , it is sometimes used in combination with the refining (or refining) process of peeling off the rice bran layer. Whitening (精白, rice whitening) or polishing (精米) refers to making white rice by removing the rice bran layer of brown rice.

Brown rice is rice in which the harvested rice is dried and threshed and then the hulls have been removed using a rubber roller machine. In addition, the area around the embryo is called the base and the opposite side is called the head, and the fine groove from the base of the surface to the head is also called the bone. The cross section of brown rice consists of the outer shell layer composed of the pericarp and seed coat from the surface, the lumen layer composed of the ectoderm and aloe layer, and the inner endosperm (starch layer). Here, the outer steel layer and the inner steel layer are collectively referred to as a rice steel layer or a steel layer.

In addition, brown rice is polished to white rice in which the rice bran layer is removed and only the starch layer is left in the milling process. Generally, milling is preferentially performed on the sides, back, and belly, and then milling is performed on the embryo and bone.

The milling degree, which indicates the degree of removal of rice bran from brown rice, has a great influence on the appearance quality, taste, and milling yield of rice.

Therefore, it is very important to mill the brown rice to an appropriate degree during the milling process. If over-refining (過搗精), that is, excessively high degree of polishing, the appearance quality of the polished rice is improved and the fatty acid value is lowered, so the deterioration due to fatty acid plaque is reduced, but the milling yield of the polished rice is decreased.

On the other hand, if the degree of milling of brown rice is low, the milling yield of white rice increases, but the taste and appearance quality are deteriorated, and the fatty acid value of brown rice increases, which worsens deterioration due to fatty acid plaque.

As described above, the degree of milling has a very important meaning in the milling of rice, but it is not clear about the milling standards related to the degree of milling.

The present invention can measure the degree of degree by measuring the color rice bran area ratio (CBB) index of the rice 7, and as a uniform light is irradiated in all directions of the rice 7 through a ring lighting, such as a shadow effect, etc. Factors that interfere with image processing can be removed, and the remaining rice bran can be measured without image loss such as color or shape of the surface of the rice grain 7 .

The present invention uses the transmitted light that transmits the rice 7, which is the measurement target, through the backlight 7 (upward illumination) of the second ring illumination lamp 15 installed under the transparent index table 4, preferably under the transparent index table 4. It is possible to detect the difference in internal density, internal cracks, and transmittance (fine granularity) of the rice 7 that cannot be seen with general visible light, and the state of the rice 7 can be detected through this.

The present invention can calculate the color rice porosity area ratio (CBB) index through the reflected image of the rice 7 and the transmitted image of the rice 7 obtained through the second camera 12, and the rice (7) through the transmitted image It is possible to detect differences in internal density, internal cracks and permeability of

The present invention can measure external characteristics, such as long and short axis, of grains to estimate the degree of sharpness and whiteness using the CBB index value and the RGB value, and to determine the cause of the milling deviation using an image processor.

In the present invention, the method for measuring the remaining rice bran is performed by the image processing unit, and by the image acquisition unit, the reflected image of the rice grain 7 placed on the transparent index table 4 is obtained using the light reflected from the ring illumination, The image acquisition unit acquires a transmission image in which the transmitted backlight passes through the rice grain 7, separates the background area and the rice field 7 area from the obtained image, and reflects the rice grain 7 After detecting the reflective residual rice bran region in the image and the transmitted residual rice bran region in the transmission image, the final residual rice bran region is detected by combining the reflective residual rice bran region and the transmitted residual rice bran region to calculate the pericarp and aloe layer area, and the calculated pericarp , it is possible to calculate the color rice bran area ratio (CBB) index of the rice (7) using the aloe layer and the starch layer.

[Measurement of Rice Quality]

Quality factors that affect the taste of rice are very diverse, and many studies have been conducted on the correlation between quality factors and taste of rice at home and abroad. Ingredients are an important factor, and after harvest, quality degradation due to enzymatic action and polishing degree, whiteness, and appearance are known as important factors.

The development of a quality measurement system that can predict the taste of rice (7) has been mainly conducted in Japan. The quality measurement system that is currently being developed and distributed is an external measurement device that measures external characteristics and a component analyzer that measures internal components. etc. Appearance measuring devices generally use an image processing system such as a camera to determine normal grains and bad grains of rice.

In the present invention, the illumination correction of the first and second cameras 11 and 12 and the infrared camera 13 is periodically performed or By setting the first and second cameras 11 and 12 and the infrared camera 13 according to the ambient illuminance atmosphere by performing the measurement at the beginning of the measurement, or when the lot of the rice 7 is changed, etc. accuracy is maximized.

In the present invention, the measured rice is discharged from the index table (4), and is applied with a roller brush (26). Repeated or continuous measurement is possible by dust collecting treatment of rice bran, fine flour, and foreign substances on the table (4) It is desirable to prevent dust and light from entering the lower part of the quality control system by dustproofing (buffering) it so that it does not affect the measurement, and covering the outside of the rice grading and quality control system with a cabinet.

The present invention measures moisture, ragi, damaged grains, heat-damaged grains, other foreign matter, and powdery grains (6 items, see Table 1) for grading of rice (polished rice) with images, and grades rice according to the government quality determination standard. It can be accurately judged as special, superior, or normal.

The present invention can determine the quality of brown rice by measuring moisture, scallops, damaged ribs, damaged ribs, and other foreign substances with an image for judging the quality of brown rice.

The present invention can measure the whiteness and milling deviation (CBB) of white rice to determine whether it exceeds or falls short of 45 degrees, which is the standard for proper milling of white rice, and can measure the milling deviation of the processed rice, that is, the degree of uniform milling.

In the present invention, the freshness of rice can be determined by predicting the germination rate by image and component measurement by a CCD camera, and component measurement can be measured using spectroscopic analysis (NIR).

Consumers of the present invention are at least 1,000 or more such as Agricultural Cooperative Rice Processing Plants (RPCs) nationwide, agricultural cooperative corporations, and rice processing companies in Korea. Rice mill).

Currently, rice grading is carried out only by sampling inspection through laboratory measuring equipment, so accurate grading of the entire rice production is impossible. In order to maintain the target yield and improve the quality of the production lot, the front and rear processing processes can be linked with the PLC to automatically control feedback.

In the present invention, rice (7) is taken as an example for grading, but in addition to rice (7), other grains such as brown rice, barley, millet, soybean, and corn are synthesized by the present invention rice grading and quality control system and method Of course, it can be determined.

The present invention described above is not limited to the present embodiment and the accompanying drawings, and various substitutions, modifications and changes are possible within the scope without departing from the technical spirit of the present invention, which is in the technical field to which the present invention pertains. It is obvious to those with ordinary knowledge.

(1)--Standing (2)--Plate
(3)--Servo motor (4)--Index table
(5)--Rice hopper (6)--Moisture meter
(7)--Rice (8)(9)--Rice supply unit
(10)--Horizontal bar (11)--First camera
(12)--Second Camera (13)--Third Camera (Infrared Camera)
(14)(15)(16)--Ring light (17)(18)(19)--Step motor
(21)(22)(23)--Color chart part (24)--Scraper
(25)--Discharge Hopper (26)--Roller Brush
(27)--Brush motor (28)--Dust collection duct
(29)--Controller (30)--Artificial Intelligence Department

Claims (10)

Plate installed on the top of the base;
an index table that is installed in the center of the plate and rotates in one direction by a servomotor;
a moisture meter installed in the rice input hopper installed on the base and measuring the moisture of the input rice and inputting it into the artificial intelligence unit;
a plurality of rice supply units installed under the rice input hopper and supplying rice to the upper surface of the index table;
a first camera installed on the base and measuring the appearance, elegance, whiteness, sage, full grain, harmless grain, and milling deviation of rice by imaging reflected light, and then inputting it into the artificial intelligence unit;
a second camera installed on the base and measuring the sarcoma by capturing the transmitted light from the rice and then inputting it to the artificial intelligence unit;
an infrared camera installed on the base and measuring the protein, amylose, fat, carbohydrate, etc. of the rice by imaging the reflected light from the rice, and then inputting it into the artificial intelligence unit;
a first ring light which is installed between the first camera and the index table and illuminates the rice;
a second ring light which is installed under the index table where the second camera is located and illuminates the rice;
a third ring lighting lamp installed between the infrared camera and the index table and illuminating the rice with infrared light;
a color chart unit that rotates by each stepper motor, moves to the lower part of the first and second cameras and the infrared camera, provides a reference illuminance for each camera, and then rotates back;
a scraper installed on the base and discharging the moving rice to the discharge hopper located at the edge of the index table;
Roller brush and brush motor installed on the base and adsorbing and cleaning rice bran, fine powder, and foreign substances on the upper surface of the index table;
a dust collecting duct installed on the upper part of the roller brush and collecting rice bran, fine powder, and foreign substances adsorbed on the roller brush;
an artificial intelligence unit connected to the controller and the controller;
Rice grading and quality management system using artificial intelligence, including The method of claim 1 , wherein:
The rice input part is disposed at the 0 ° position of the index table, and counterclockwise based on this, the first camera is at a position of 60 °, the second camera is at a position of 120 °, an infrared camera is located at a position of 180 °, and a scraper is located at a position of 240 °, Rice grading and quality control system using artificial intelligence, characterized in that a roller brush is placed at the 300° position. 3. The method according to claim 1 or 2:
The artificial intelligence unit,
a data collection unit for collecting rice moisture data input through a moisture meter and rice image data input from the first and second cameras and infrared cameras;
a database for storing collected data;
a data processing unit that processes the collected data;
artificial intelligence learning unit that learns from processed data;
a determination unit that determines a rice grade based on the learned data, and presents the whiteness, milling deviation, and component data of the rice;
Rice grading and quality management system using artificial intelligence, including 3. The method of claim 1 or 2:
a horizontal bar that evenly spreads the rice supplied to the index table on the upper surface of the index table and maintains it as one layer;
Rice grading and quality management system using artificial intelligence further comprising a. 3. The method of claim 1 or 2:
The first and second cameras and the infrared camera are CCD cameras, and the rice grain grading and quality management system using artificial intelligence, characterized in that it includes an image processing module and an automatic color & illumination correction algorithm. 3. The method of claim 1 or 2:
The first, second, and third ring lights are provided with a ring-shaped PCB on which a plurality of LEDs are mounted at predetermined intervals on the inner inclined surface of a circular ring having a through hole in the center, and the illumination light is focused toward a point on the central virtual line of the through hole. Rice grading and quality management system using artificial intelligence, characterized in that it has a predetermined slope so that it can be 3. The method of claim 1 or 2:
The first and third ring lights are installed on the index table so that the reflected light of the rice is incident on the first camera and the infrared camera, respectively, and the second ring lights are installed on the lower part of the index table so that the transmitted light of the rice is incident on the second camera Rice grading and quality management system using artificial intelligence, characterized by a) the color chart unit rotates at right angles by the stepping motor and moves to the lower portions of the first, second and infrared cameras, respectively;
b) the first, second and infrared cameras photograph the surface of the color chart and then calibrate the illuminance using an automatic color & illuminance correction algorithm;
c) returning the color chart unit to a position outside the index table by the stepping motor;
d) inputting the moisture data measured by the moisture meter on the rice fed into the rice hopper to the artificial intelligence unit;
e) supplying the rice from the rice input hopper to the upper surface of the index table rotating in one direction through the rice input unit;
f) taking the reflected light of the rice moved to the lower part of the first camera, measuring the appearance, dignity, whiteness, sage, perfect rib, harmless rib, and grain deviation, etc., and inputting the image to the artificial intelligence unit;
g) measuring the serous granularity by imaging the transmitted light of the rice that has moved to the lower part of the second camera, and then inputting it to the artificial intelligence unit;
h) measuring the protein, amylose, fat, carbohydrate, etc. of the rice by imaging the infrared reflected light of the rice moved to the lower part of the infrared camera and inputting it to the artificial intelligence unit;
i) dropping and discharging the rice moved to the scraper into the discharge hopper at the edge of the index table;
j) adsorbing fine powder, rice bran, foreign matter, etc. remaining on the index table surface with a roller brush and then collecting dust with a dust collecting duct on the top of the roller brush; and
k) the artificial intelligence unit learning the data input from the moisture meter and the first, second, and third cameras to determine the rice grade as special, prize, normal, etc.;
Rice grading and quality control method using artificial intelligence, including 9. The method of claim 8;
l) the artificial intelligence unit presenting the whiteness, grain deviation, and component data of the rice;
Rice grading and quality control method using artificial intelligence further comprising a. 10. The method of claim 8 or 9:
In step e) in which the rice of the rice input hopper is supplied to the upper surface of the index table rotating in one direction through the rice input unit, the rice supplied to the index table is evenly spread by the horizontal bar and moving along the index table ;
Rice grading and quality control method using artificial intelligence further comprising a.

Images