KR20090112421A - System and method for evaluating the grade of rice grains - Google Patents

Abstract

PURPOSE: A system and method for evaluating a grade of rice grains are provided to evaluate not only shape and condition but also color and brightness. CONSTITUTION: A sample input part(110) inputs a sample to be classified. A conveying/classifying part(120) transfers while classifying each grain of the sample supplied through the sample input part. A light source part irradiates the light to the sample classified by the conveying/classifying part. An optical detecting part(160) detects the transmission light permeating the sample and reflection light. A management system(170) takes an image of the sample and distinguishes a grad of the sample based on the amount of the transmission light and reflection light.

Description

Particle quality discrimination system and its method {SYSTEM AND METHOD FOR EVALUATING THE GRADE OF RICE GRAINS}

The present invention relates to a system for determining a fine grain quality and a method thereof, and more particularly, in determining a fine grain quality, an abnormal state of a shape (shape) that may affect quality as well as identification due to differences in color and brightness. A fine quality identification system and its quality that can be identified and selected quickly and accurately, can be quickly and accurately determined by using image processing technology, and can be identified by checking the image by humans rather than by conventional mechanical or non-objective determination. It is about a method.

Today, the amount of rice (grain) imported from the expansion of the rice market opening under the World Trade Organization (WTO) system is increasing, which threatens the livelihood of agriculture / farming.

In response to this, in order to secure the competitiveness of the rice crop industry, measures should be taken to promote the supplementation of the rice industry to stabilize farm incomes in connection with the agricultural / farmland comprehensive measures.

Accordingly, a plan to promote rice's quality competitiveness has been proposed as a brand development plan.The Ministry of Agriculture, Farming Organizations, Nonghyup and Rice Process Complex (RPC) will develop high-quality rice brands, varieties and After establishing the rice brand by differentiating the cultivation method, the plan is to support the facility management and funding for the excellent brand management body.

Such a rice brand can be differentiated from the existing rice in the market, if the consumer's trust is obtained, not only can stabilize the income of farmers, but can also consume the rice produced in agriculture / rural land in Korea.

In order to foster the rice brand, quality control of brand rice should be executed first, and for quality control, development of a quality discrimination system capable of judging the quality of rice should be made first.

In the past, when purchasing rice produced by individual farmers, the price of rice to be purchased is accurately determined, that is, the amount of perfect and bad rice is accurately measured, and the purchase price is determined according to the degree of rice production to rice production. Impartiality between buyers and buyers.

Such determination of the quality of the rice requires an objective and accurate determination since it causes problems such as a decrease in rice production per rice yield due to poor management after harvesting if the determination of the quality of the rice is not accurate.

To determine the quality of rice, the agricultural inspector collects a sample by using a color band (Tae-gwan; an extension to stab and remove grains in the bale), and visually evaluates the grains with the naked eye, a moisture meter, and a weaving machine (moisture content, volume weight). , Degree, foreign matter and damage grains, etc.).

As described above, the work of determining the rice quality of the conventional rice has a problem that it takes a lot of time with a high-quality manpower (agricultural inspector who is well-versed in grain inspection) and difficult to maintain fairness and objectivity in the quality determination work.

In order to solve such a problem, the apparatus which irradiates light to a sample, detects the quantity of light reflected or transmitted, and determines the grade was developed and introduced.

However, since the existing class determination apparatus imports a foreign product and is used in Korea as it is, an error occurs in the class determination due to the difference in grain size of rice grains caused by the difference in the environment between foreign countries and domestic countries.

In addition, since the rice brand must be operated as a private company, it is difficult to employ a quality control agent (agricultural inspector) in a private company.

In addition, the existing quality determination device is provided with a groove (cavity) in the tray (tray) for transporting the sample of fine grain, and adopts a method of transferring the sample through the groove, which is caused by the difference in the environment between foreign and domestic An error occurs due to the difference in the grain size (grain size) of the grain, or the grade cannot be accurately determined.

Therefore, there is no need for a separate quality control personnel to fit the domestic environment, and the development of a quality determination system capable of accurately determining the quality is urgently needed.

The present invention has been proposed to meet the necessity as described above, and an object thereof is to provide a fine grain quality discriminating system and a method for quickly and accurately determining the grain quality of rice grains grown in a domestic environment.

In addition, the present invention, when determining the fine particles, fine particle identification system that can quickly and accurately identify / select the abnormal state of the shape (shape) that may affect the quality as well as the identification by the difference in color and brightness And a method thereof.

In addition, the present invention provides a fine grain quality discrimination system and method that can be quickly and accurately discriminated by the use of an image processing technology, and the management personnel can discriminate the quality of the human beings rather than the existing mechanical or non-objective discrimination. Its purpose is to.

Particle quality discrimination system according to an aspect of the present invention, a sample inlet for injecting a sample to determine the grade, and a sample transfer / separation unit for separating each grain of the sample while transferring the sample introduced through the sample inlet; And a light source unit for irradiating light to the sample from which the grain is separated in the sample transfer / separation unit, a light detector for detecting the transmitted light reflected from the light source unit and the reflected light reflected from the light source unit, and the light detector. And a management system that acquires an image of the sample based on the amount of transmitted light and reflected light, and processes the image to determine the quality of the sample.

The sample transfer / separator may be configured to separate the sample transfer unit for transferring to the discharge port while passing between the light source unit and the light detection unit facing the input sample, and to prevent each grain of the sample transferred through the sample transfer unit from contacting each other. A brush, the brush is attached to the outer peripheral surface, and a roller for rotating the brush, and a motor for providing a rotational force to the roller.

In the fine granularity determination system, the brush separates the grain while performing a constant velocity or acceleration / deceleration rotational forward movement, and the sample transfer / separation unit guides the forward movement of the brush and is located at both end sides of the roller. It further comprises two guide rails.

The brush has a distance between the ends of neighboring brushes set to a predetermined length longer than the average length or the average length of the grains, arranged side by side with neighboring brushes, or the end forming shapes of surrounding brushes arranged in a lattice manner. Particle grade discrimination system, characterized in that.

The fine grain discrimination system further includes a transfer control unit for controlling driving of the motor in response to an input of a management personnel while supplying driving power to the sample transfer / separation unit, wherein the light source unit is in the same direction as the light source detection unit. And a first light source unit for irradiating light with a sample, and a second light source unit facing each other with the light source unit and the sample therebetween and illuminating light in the direction of the sample.

The light detector detects an amount of reflected light that is irradiated from the first light source and reflected on the sample and an amount of transmitted light that is irradiated from the second light source and transmitted through the sample.

The management system includes a power supply unit for supplying driving power to the light source unit and the sample transfer / separation unit, an input unit for receiving input information from a management agent, a display unit for outputting a display screen, and input information input through the input unit. According to the main control unit for controlling the driving of the light source unit and the sample transport / separation unit, and based on the amount of reflected light and transmitted light received from the light detection unit to obtain an image of each grain of the sample, and the operation based on the image to process the sample It includes an analysis unit for determining the quality of the.

The analyzing unit converts the transmitted light received from the light detector into a gray level, applies an anisotropic diffusion filter, converts the gray level image into a binarized image, and converts the binary inverse, and then numbers each grain. The shape information of each grain is extracted, and the grade is discriminated based on the grade determination criteria.

The analysis unit distinguishes each grain based on two points where corner points are closest to each other by using Harris corner detection in the binary inversely transformed image, and arranges a blur value generated at the edge of the grain. And inputting the transmitted light and the reflected light image of each grain of the sample, the number information and each grain value assigned to each grain, and storing the image of each grain based on the extracted position information of each grain. After calculating the center point of each grain and the length of the horizontal / vertical, and calculating the area of each grain in pixel units, the circularity is calculated to extract the shape information.

The analysis unit calculates the circularity by using Equation 1 below, and the closer to the circle, the closer the value is to 1.0.

[Equation 1]

The light source unit irradiates medium infrared rays or near infrared rays to the sample in the particulate matter discrimination system, and the light detector detects wavelengths of the medium infrared rays or near infrared rays passing through the sample.

The management system analyzes qualitative and quantitatively by inspecting atomic spectra based on the wavelength of the medium or near infrared rays detected by the light detection unit.

According to another aspect of the present invention, a quantitative analysis system of rice grains using a spectroscopic analysis method includes: a light source unit for irradiating medium or near infrared rays with a sample, and a medium or near infrared ray irradiated from the light source unit and passing through the sample; And a light detector and a analyzer for qualitative analysis based on the middle infrared rays detected by the light detector, quantitative analysis based on near infrared rays, and correcting / modeling qualitative and quantitative values.

According to still another aspect of the present invention, there is provided a method for determining a particulate quality, comprising: injecting a sample into a mirin quality determination system, irradiating light to the injected sample, and transmitting and reflecting reflected light through the sample. Detecting, obtaining shape information of an image based on the transmitted light and reflected light, and determining the quality of the sample based on the image.

The shape information of the image may include: an image acquiring step of acquiring an image based on the amount of reflected light and the amount of transmitted light, a preprocessing step of converting the transmitted light into a gray level, and applying an anisotropic diffusion filter; A segmentation step of converting an image of the image into a binarization image, a binary inverse transformation step of interchanging a background and a grain divided in the binarization image, a labeling step of numbering each grain, and shape information of each grain Shape information extraction step of extracting.

The granularity discrimination method includes an edge division step of distinguishing each grain based on two points where corner points are closest to each other by using Harris corner detection, and a blurring value generated at an image border of each grain. It further includes a border cleanup step to clean up.

The extracting of the shape information comprises: receiving transmitted and reflected light images for each grain, number information assigned to each grain and each grain boundary value; extracting position information for each grain number; Storing each grain as an image based on the work position and end position of each grain, calculating the center point of each grain and the length of the horizontal / vertical length, and calculating the area of each grain in pixel units. And calculating a circumferential length based on the length of the outline of each grain, and calculating the circularity of each grain.

The irradiating of the light may include generating reflected light by irradiating first light in the same direction as the light detector that detects the light, and generating second light to face the light detector and the sample therebetween. It is characterized by generating the transmitted light.

The method of determining the fine grain quality, the step of irradiating the sample material with medium or near infrared rays, the step of detecting the medium or near infrared light passing through the sample, qualitative analysis based on the medium infrared rays, Quantitatively analyzing based on the near infrared ray, and correcting / modeling the numerical value of the qualitative and quantitative analysis results.

According to the present invention as described above, it is possible to quickly and accurately identify / select the abnormal state of the shape (shape) that may affect the quality as well as the identification by the difference in color and brightness when determining the particulate quality. .

In addition, according to the present invention, the image processing technology can be used to quickly and accurately discriminate, and the identification process can be performed while the human being, that is, the management personnel, checks the processed image rather than the existing mechanical or non-objective determination.

In addition, according to the present invention, various data analysis techniques such as average, minimum, maximum, standard horseshoe, and frequency can be applied.

Therefore, consumers can be trusted for the quality of rice, and producers can manage the quality without consuming much manpower and cost for quality control, so that high added value can be realized, and management personnel are simple / easy, Grade can be determined quickly and accurately.

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS Hereinafter, a fine grain quality discriminating system and a method thereof according to a preferred embodiment of the present invention will be described in detail with reference to the accompanying drawings.

In the following detailed description of the present invention, for example, a case of determining the quality of rice (rice) is described, but other grains (rice), grains (barley, wheat, corn, etc.), special crops (sesame, perilla, etc.) The same applies to the case of determining the elegance of.

1 is a block diagram illustrating a granular product discrimination system according to a preferred embodiment of the present invention.

Referring to FIG. 1, the particulate grade determination system 100 according to the present invention includes a sample input unit 110, a light source unit 140, a light control unit 150, a sample transfer / separation unit 120, and a transfer control unit 130. ), The light detector 160, and the management system 170.

The sample input unit 110 transmits (discharges) the sample transfer / separation unit 120 when a management agent who determines / manages the quality of the fine particles puts the sample into the input port (not shown).

The sample transfer / separator 120 separates each sample (kernel) while transferring a sample delivered through the sample input unit 110. That is, the sample transfer / separation unit 120 separates the grains (k of FIG. 4) so that they do not come into contact with each other.

The transfer control unit 130 supplies driving power to the sample transfer / separation unit 120 and controls the transfer of the sample.

Details of the sample transfer / separator 120 and the transfer control unit 130 will be described later.

The light source unit 140 may be implemented as a light emitting device such as a light emitting diode (LED). When the power is supplied, the light source unit 140 irradiates light toward the sample.

The light control unit 150 supplies driving power to the light source unit 140, controls the supply of the driving power, and controls ON / OFF of the light source unit 140.

The light detector 160 may be implemented as a means for detecting light, such as a camera. The light detector 160 detects (measures) the light transmitted from the light source unit 140 and the reflected light reflected by the sample. The light detector 160 transmits the detected amount of transmitted light and the amount of reflected light to the management system 170.

The management system 170 controls the light emission of the light source unit 140 through the light control unit 150 according to the input of the management personnel, and controls the sample transfer / separation unit 120 through the transfer control unit 130.

In addition, the management system 170 determines a corresponding particulate quality based on the amount of transmitted light and the amount of reflected light received from the light detector 160.

2 is a view for explaining a sample transfer / separation unit according to the present invention.

Referring to FIG. 2, when the sample discharged from the sample input unit 110 is seated, the sample transfer / separation unit 120 moves the sample to be passed between the light source unit 140 and the light detection unit 160. A sample feeder 121, a motor 122 that generates rotational force by the drive power provided from the transfer control unit 130, a roller 123 that rotates by the rotational force provided from the motor 122, It is attached to the outer circumferential surface of the roller 123, and includes a plurality of brushes 124 to separate each sample so as not to contact.

The sample transfer unit 121 may be embodied in a conveyor manner, and the sample transfer unit 121 may be transferred to a sample discharge port (not shown) separately provided while passing through the light source unit 140 and the light detector 160. In addition, the seating table 126 provided on the sample transfer part 121 is mounted with a sample is preferably made of a transparent material so that the light irradiated from the light source unit 140 is transmitted through the sample.

 The transfer control unit 130 may include a first power supply unit 131 for providing driving power to the motor 122, and a first control unit 132 for controlling rotational force of the motor 122 and sample transfer of the sample transfer unit 121. It includes.

The transfer control unit 130 transfers the sample on which the sample transfer unit 121 is seated according to the transfer control of the management system 170, and supplies driving power to the motor 122 to provide rotational force to the roller 123, or The driving of the sample transfer unit 121 and the motor 122 is stopped.

That is, the transfer control unit 130 drives the motor 122 and the sample transfer unit 121 during the fine grade determination operation according to the input of the management personnel, and when the grade determination operation is completed, the motor 122 and the sample transfer unit ( 121) is stopped.

That is, the constant velocity or acceleration / deceleration rotational forward movement of the brush 124 is controlled by the feed control unit 130 controlling the rotational force of the motor 122, and the guide role of the brush 124 is the two sides of the brush 124, that is, A pair of guide rails 125 located at both end sides of the roller 123 processes.

3 is a view for explaining the sample transfer unit according to the present invention in more detail, as shown in Figure 3, a pair of guide rails 125 located on both sides of the upper portion of the sample transfer linear movement of the roller 123 That is, the linear motion of the brush 124 is guided.

In addition, the constant speed or acceleration / deceleration rotation of the brush 124 may be determined according to the amount of sample and the type of sample to be injected, and the management personnel may experimentally derive a stable speed and acceleration.

The first control unit 132 of the transfer control unit 130 controls the ON / OFF by supplying / stopping the driving power supplied to the motor 122 according to the motor 122 control signal received from the management system 170, PID (proportional, integral and differential) control of the rotational force of 122 controls the constant velocity or acceleration / deceleration rotational forward motion of the brush 124.

In addition, each brush 124 attached to the outer circumferential surface of the roller 123 is contacted with a sample, ie, a grain k, seated on the seat 126 of the sample transfer part 121 by the rotation of the roller 123. Do not separate it.

Therefore, rather than providing a groove in the conventional tray (mounting table 126), since each grain (k) of the sample is separated through the brush 124, due to the difference in the grain size (k) of the sample The error of the discrimination can be prevented in advance, and the grade can be discriminated more accurately.

4 is a view for explaining a method for separating the sample transport / separation unit according to the present invention.

As shown in FIG. 4, the sample seated on the seating table 126 may be introduced in a unit bundle from the sample input unit 110, so that the sample, that is, the grain k, is bundled. In this way, when the grains k are aggregated, each grain k should be separated from each other so that the grains k do not come into contact with each other because the transmission and reflection of light are not precisely performed for each grain k.

Therefore, the brush 124 attached to the outer circumferential surface of the roller 123 is separated so that the grains k do not come into contact with each other by the rotation of the roller 123. At this time, the distance (a) between the ends of the neighboring brush 124 is set longer than the average length or the average length of the corresponding grain (k), and only one grain (k) is accommodated between the brushes (124) It is desirable to allow each grain k to be separated.

5A and 5B are views for explaining a brush arrangement of a sample transfer / separator according to a preferred embodiment of the present invention.

5A and 5B illustrate a state in which the brush 124 attached to the outer circumferential surface of the roller 123 is unfolded on a flat surface for convenience of description, and as shown in FIG. 5A, the distance a between the ends of the brushes 124 is Constant, each brush 124 may be arranged side by side.

In addition, as shown in Figure 5b, the brush 124 is implemented in the form of a piece rather than a date, arranged in a lattice manner, the grain (k) of the sample between the neighboring brush 124 before / after and left / right Can be converged.

At this time, the distance (a) between the ends of the brush 124 is preferably set to the average size of the grain (k) of domestic rice.

6 is a view for explaining a light source unit and a light detector according to a preferred embodiment of the present invention.

Referring to FIG. 6, the first light source unit 140-1 and the second light source unit 140-2, which are seated on the mounting table 126 and face the grain k of each sample separated by the brush 124, Irradiate light.

In addition, the light detector 160 may be implemented as an optical measuring system. The light detector 160 detects the reflected light reflected by the sample from the light emitted from the first light source 140-1 and is irradiated from the second light source 140-2. The transmitted light passing through the sample is detected, and the amount of reflected light and the amount of transmitted light are transmitted to the management system 170. In this case, the first light source unit 140-1 and the second light source unit 140-2 may irradiate light having different wavelengths, so that the light detector 160 may distinguish the reflected light and the transmitted light.

7A and 7B are diagrams for describing a method of detecting reflected light and transmitted light.

As shown in FIG. 7A, when the first light source unit 140-1 irradiates the sample with light in the same direction as the light detector 160, the sample reflects the irradiated light, and the light detector 160 The light reflected by the grain k of the sample is detected.

As shown in FIG. 7B, when the second light source unit 140-2 irradiates the tube with the sample in the direction of the sample different from the light detector 160, the sample transmits light according to the transmittance and the light detector 160 detects the transmitted light transmitted through the grain (k) of the sample.

The light control unit 150 includes a driving power supply unit 151 and a second control unit 152 for supplying driving power to the first light source unit 140-1 and the second light source unit 140-2. 152 supplies / stops the driving power applied to the first light source unit 140-1 and the second light source unit 140-2 according to the light control signal received from the management system 170.

8 is a block diagram illustrating a management system according to a preferred embodiment of the present invention.

Referring to FIG. 8, the management system 170 according to the present invention includes a power supply unit that provides driving power to the sample transfer / separation unit 120 and the light source unit 140 through the transfer control unit 130 and the light control unit 150. 173, the light control signal to the light control unit 150 according to the input unit 171 for inputting information from the management personnel, the display unit 172 for outputting the display screen, and the input information input through the input unit 171. The main control unit 174 which transmits the control signal of the motor 122 to the transfer control unit 130 and controls the overall operation of the fine grain discrimination system 100, and the reflected light and the transmitted light received from the light detection unit 160. It includes an analysis unit 175 for processing an analysis operation for determining the fine quality according to the amount of light.

The main control unit 174 transmits a motor 122 control signal to the transfer control unit 130 and transmits a light on control signal to the light control unit 150 when the management agent inputs a start of work through the input unit 171. When the work of determining the quality is started, and when the work is finished or the work is interrupted, the motor 122 control signal is transmitted to the movement controller, and the light control signal is transmitted to the light controller 150 to perform the work of quality determination. Allow it to end or stop.

The input unit 171 may be implemented as a keypad or a touch screen method having a plurality of keys, and transmits input information input from a management personnel to the main control unit 174.

The analyzer 175 analyzes / determines the quality of the sample based on the amount of reflected light and the amount of transmitted light received from the light detector 160.

Hereinafter, a method of analyzing the sample (k) by the analyzing unit 175 will be described in detail with reference to FIGS. 9A to 9H.

9A to 9H are views illustrating an image according to a process of analyzing a sample by an analysis unit according to a preferred embodiment of the present invention, and FIG. 10 is a flowchart illustrating a method of analyzing a sample according to the present invention.

The analysis unit 175 of the management system 170 performs an image acquisition process of acquiring an image based on the amount of reflected light and the amount of transmitted light received from the light detector 160 (S 100).

The analyzer 175 converts the transmitted light received from the light detector 160 into a gray level to obtain an image as shown in FIG. 9A, and applies a anisotropic diffusion filter to obtain an image as shown in FIG. 9B. Perform (S 110). At this time, the boiling cushion diffusion filter may be applied a plurality of times, for example five times.

The analyzer 175 performs an area segmentation process of obtaining a binarized image as illustrated in FIG. 9C from the gray level image obtained as illustrated in FIG. 9B (S120). In this case, the analyzer 175 may acquire the binarization image by reflecting the binarization threshold value in the gray level, and the binarization threshold value may be determined based on the mode method.

In addition, the analysis unit 175 has a zero value (black) of the sample divided in the binarization image of FIG. 9C and a background value of 255 (white). A binary inverse transformation process is performed to represent the white color (S 130).

On the other hand, the analysis unit 175 may be in contact with some of each grain (k) mechanically separated by the brush 124 of the sample transfer / separation unit 120, the contacted grain (k), That is, it distinguishes the attached grains (k). At this time, the analysis unit 175 performs an edge division process for distinguishing each grain k based on two points where corner points are closest to each other using Harris corner detection (S 140).

And, the analysis unit 175 cleans up the edges to clean up the interference value between the adjacent images, the garbage value of the edge generated in the preprocessing process (S 110), that is, the blurry value generated at the edge of the grain (k). Perform the process (S 150),

The analysis unit 175 performs a labeling process of assigning a number to each grain (k) in the image of the grain (k) sharpened in the edge cleanup process (S 160).

Then, the analysis unit 175 performs a shape information extraction process for extracting the shape information of each grain (k) (S 170).

11 is a flowchart for explaining a shape information extraction process according to the present invention.

Referring to FIG. 11, the analyzer 175 receives and sets transmitted and reflected light images for each grain k, number information assigned to each grain k, and each grain boundary value as shown in FIG. 9G (S). 171).

Table 1 below is for explaining each threshold applied to the present invention.

Threshold Type Explanation One Actual average length of grain (k) (mm) Define the size of the grain (k) of the actual sample, and define the size of the grain (k) in the image 2 Actual average length of the grain (k) Calculate the average length of the full grain long axis of the image of the grain (k) and fertilize with the actual average length (mm) to calculate the actual length per pill 3 Average area of grain (pixel) Calculate the complete grain area in pixels, and use it to determine damage and compare damage area of damage grain 4 Fertilizer area (% / 100) of damage grain (k) Defining ratios to determine by damage grain 5 Damage area (k) area standard (pixel) Pixel value to turn into damaged grain when red or black tide is over a certain part of total area 6 Min_ratio Minimum percentage value of complete grain, for distinguishing heterogeneous grains 7 Max_ratio A value for distinguishing heterogeneous grains by the maximum ratio value of perfect grains 8 Corner finder 1 Recognize the boundary only when the relative positions of the retrieved corners are less than or equal to the first corner value 9 Corner find 2 Threshold that is referred to the Harris corner find algorithm, the criterion for recognizing corner strength (higher values yield fewer corners, lower values search more konas).

The analysis unit 175 extracts the position information of the grain k (S 172). At this time, the analysis unit 175 may detect the start and end positions of the number k given each number based on the number information, the analysis unit 175 detects the start and position information for each number.

The analyzer 175 stores each grain k as an image based on the start and end positions of each grain k (S 173). The image of each grain k is stored as shown in FIG. 9G.

The analyzing unit 175 finds the center point of each grain k (S 174), and calculates lengths of the width and length of each grain k (S 175). The analysis unit 175 determines the horizontal value of the lightest large value passing through the center point in the image of the grain k from 0 degrees to 360 degrees based on the center point of each grain k, and the horizontal value. The length of the straight line passing through the center point plus 90 degrees is determined as the vertical value.

The analysis unit 175 calculates the area of each grain in pixel units (S 176), finds an outline in the image of each grain, and finds the length of the outline to calculate the perimeter length. (S 177). The calculation of the circumference length is calculated by the straight line as 1 and the diagonal as sqrt (2) (square root (2)).

The analyzer 175 may extract the edge of each grain k to calculate the circumferential length and circularity of each grain k as shown in FIG. 9H.

The analysis unit 175 calculates a circularity which is a complexity of the image shape of the grains k based on the area and the perimeter length of each grain (S 178).

The analysis unit 175 calculates the circularity as shown in Equation 1 below.

는 원형도를 의미하며, 반경 r인 원의 경우, 둘레 길이가 2πr, 면적 π

이므로,

=1이 된다. 따라서, 낱알(k)의 오브젝트가 원형에 가까울수록

값이 1.0에 근접하고, 복잡할수록 1.0에서 작아지는 값을 가지게 된다.In Equation 1

Means circularity, for a circle with radius r, the perimeter is 2πr, area π

Because of,

= 1. Therefore, the closer the object of grain (k) is to the circle,

The closer the value is to 1.0, the smaller the value becomes.

On the other hand, when the shape information extraction process is completed, the analysis unit 175, based on the received threshold value, the image of the extracted grain (k) (reflected light, transmitted light), area, circumferential length, circularity, etc. Determine the grade of (S 180).

The criteria for determining the quality of the analysis unit 175 is based on the quality determination host of the agriculture and forestry department, for example, can be determined as follows.

12A to 12G are diagrams for explaining an example of determining the quality of a sample by the particulate quality determining system according to the present invention. The left image of FIG. 12 is a transmitted light image, and the speculative image is a reflected light image.

If the analysis unit 175 is an image of a grain (k) as shown in FIG. 12A, the damage ratio with respect to the total area is less than the comparison area (% / 100) of the damage grain (k) due to red tide. It is determined by red tide because it is larger than the area reference (pixel).

In the case of FIG. 12B, the damage ratio is determined as the damage grain because it is larger than the comparative area (% / 100) of the damage grain (k). In the case of FIG. 12C, the fraction of powdery quality for the total area is larger than the threshold value. It is discriminated by a fine grain (cloudy grain (k)).

In the case of FIG. 12D, the damage ratio of the total area is smaller than the comparison area (% / 100) of the damage grain (k), but is larger than the area reference (pixel) of the damage grain (k). 12E, the length of the long axis of the grain k is smaller than the number 1 body for filtering the grain k, and is determined as a wrap. In the case of FIG. 12F, the top axis is determined based on the short axis direction of the grain k. Since the difference between the brightness of the part and the lower part is obvious, it is determined by the copper grain (gold grain (k)), and in the case of FIG. 12G, the average length (pixel) and the area (pixel) of the grain (k) are satisfied. Since it is smaller than the fertilizer area of rice grains (% / 100) and the area of damage grains (k), it is determined as a complete grain.

That is, the analysis unit 175 recognizes shapes such as the area, length, width, cross / vertical ratio of the grains k, based on the images of the grains k of the sample, and identifies the pigmentation and the partial pigmentation. Recognize the distribution and pattern of color for discriminating upright, unestablished, and damaged grains, and recognize the change of concentration to distinguish Dongha, partial coloring, germination, and germ granules, and reflect them in measurement index for quality evaluation To determine the grade.

Measurement indicators for the evaluation of quality include the shape of the grains (k) such as the number, length, width, thickness, area, aspect ratio, perimeter, etc. Based on the state of the grain (k), such as small flies, foreign matters, nuclei, heterogeneous grains, and donghalib.

13 is a view showing a display screen for determining the quality of the sample in the particulate quality determination system according to the present invention.

FIG. 13 is a screen for outputting the display unit 172 of the management system 170 to be confirmed by the management personnel. The transmitted light and the reflected light image of each grain k, the threshold values and the respective grains k of the sample. The information for determining the grades, ie, perfect grain, damage boundary, large flies, small flies, foreign objects, nucleus, double grains, copper grains, and degree, are displayed as a percentage.

Therefore, the management personnel who are not experts in determining the fine quality can manage the quality of the sample.

On the other hand, the granular grade determination system 100 according to the present invention can not only determine the grade by processing the grain (k) image of the sample, but also can analyze the qualitative and quantitative quantity of the sample.

14 is a block diagram illustrating a quantitative analysis method of rice grains using spectroscopic analysis according to a preferred embodiment of the present invention.

Referring to FIG. 14, the light source unit 140 of the present invention irradiates near-infrared and mid-infrared rays with a measurement grain, that is, a sample, and the light detector 160 detects near-infrared and mid-infrared rays passing through the sample to manage the system. And transmits the data to the analysis unit 175 of 170.

The analyzer 175 analyzes near and mid-infrared rays detected by penetrating the sample, and performs qualitative and quantitative analysis of rice grains.

15 is a flowchart for explaining a method of quantitative analysis of rice grains using spectroscopic analysis according to a preferred embodiment of the present invention.

Referring to FIG. 15, the main control unit of the management system 170 controls the wavelength of the light generated through the light control unit 150 so that mid-infrared (MIR) in qualitative analysis and near-infrared (NIP) in quantitative analysis are irradiated. (S 200).

Infrared rays can be classified into Near IR (NIR) in the 700 ~ 3000 nm range and MIR Medium IR in the 3 ~ 30 micrometer (1 um = 1000 nm) range. -CH present in the infrared. The basic molecular kinetic energy of the OH and _NH bonds appears in the combination and overtone bands of the energy, and the absorbance is alleviated in the near infrared, which appears in the bond and overtone bands.

Therefore, the main control unit 174 causes the infrared light to be irradiated from the light source unit 140 during qualitative analysis, and the analysis unit 175 analyzes the quality of the sample based on the light detected by the light detector 160. Component) is analyzed (S 210).

For example, the analysis unit 175 analyzes the components constituting the grain (k) of the material, that is, components such as water, protein, lipids, fiber, carbohydrate and the like.

In addition, the main controller 174 is a qualitative analysis is completed so that near infrared light is irradiated from the light source unit 140 during the quantitative analysis, the analysis unit 175 quantitates the sample based on the light detected through the light detector 160 (Quantitative relationship between components) is analyzed (S 220).

At this time, the analysis unit 175 qualitatively and quantitatively analyzes the atomic spectra of the near infrared and the middle infrared rays passing through the sample.

In addition, the analysis unit 175 corrects and models numerical values of the quantitative relationship between the components of the sample and each component by a regression analysis technique (S 230), and verifies and predicts the components of the sample (S 240). ).

As such, the qualitative and quantitative analysis method using spectroscopic analysis using near-infrared and mid-infrared rays is a non-destructive analysis method, so that the sample is not destroyed, no sample preparation is required, and various components of the sample can be analyzed simultaneously. Quantitative and confirmatory tests and the like can be processed, and the physical quantity measuring device is provided on the seating table 126 on which the sample is seated, so that the physical quantity can be simultaneously measured.

In addition to solid samples, samples such as liquids, powders, suspensions, gels, creams, and solutions can be analyzed, quantitatively quantitatively analyzed, and there is no limit to the container that holds the samples, and can be quantified from tens of ppm to tens of percent. Analysis is possible.

Although the present invention has been described in detail only with respect to the described embodiments, it will be apparent to those skilled in the art that various modifications and changes are possible within the technical spirit of the present invention, and such modifications and modifications belong to the appended claims.

BRIEF DESCRIPTION OF THE DRAWINGS Fig. 1 is a block diagram for explaining a granularity discrimination system according to a preferred embodiment of the present invention.

Figure 2 is a view for explaining a sample transfer / separation unit according to the present invention.

3 is a view for explaining a sample transfer unit according to the present invention in more detail.

4 is a view for explaining a method for separating the sample transport / separation unit according to the present invention.

5a and 5b is a view for explaining the brush arrangement of the sample transfer / separation unit according to a preferred embodiment of the present invention.

6 is a view for explaining a light source unit and a light detection unit according to a preferred embodiment of the present invention.

7A and 7B are diagrams for explaining a method of detecting reflected light and transmitted light.

8 is a block diagram illustrating a management system according to a preferred embodiment of the present invention.

9A to 9H are views illustrating an image according to a process of analyzing a sample by an analysis unit according to a preferred embodiment of the present invention.

10 is a flowchart for explaining a method for analyzing a sample according to the present invention.

11 is a flowchart for explaining a shape information extraction process according to the present invention.

12A to 12G are diagrams for explaining an example of determining the quality of a sample by the particulate quality discriminating system according to the present invention.

13 is a view showing a display screen for determining the quality of the sample in the particulate quality determination system according to the present invention.

14 is a block diagram illustrating a quantitative analysis method of rice grains using spectroscopic analysis according to a preferred embodiment of the present invention.

15 is a flowchart for explaining a method for quantitative analysis of rice using spectroscopic analysis according to a preferred embodiment of the present invention.

<Description of the symbols for the main parts of the drawings>

100: fine grain discrimination system 110: sample input unit

120: sample transfer / separation unit 121: sample transfer unit

122: motor 123: roller

124: brush 130: transfer control

131: power supply unit 132: first control unit

125: guide rail 126: seating table

140: light source unit 150: light control unit

151: driving power supply unit 152: second control unit

160: light detector 170: management system

171: input unit 172: display unit

173: power supply unit 174: main control unit

175: analysis unit

Claims (21)

In a grain grade discrimination system, A sample inlet for injecting a sample for quality determination; A sample transfer / separation unit for separating each grain of the sample while transferring the sample introduced through the sample inlet; A light source unit for irradiating light to the sample from which the grains are separated from the sample transfer / separation unit; A light detector for detecting transmitted light reflected from the light source unit and transmitted reflected light through the sample; And a management system for acquiring an image of the sample based on the amount of transmitted and reflected light detected by the light detector and processing the image to determine the quality of the sample. The method of claim 1, wherein the sample transfer / separation unit, A sample transfer unit configured to transfer the discharged sample to the discharge port while passing between the light source unit and the light detection unit facing the input sample; A brush that separates each grain of the sample transferred through the sample transfer unit from being in contact with each other; The brush is attached to the outer peripheral surface, the roller for rotating the brush, Particle quality determination system having a motor for providing a rotational force to the roller. The method of claim 2, The brush separates the grain while making constant or acceleration / deceleration rotational forward movement, The sample transfer / separation unit, And a plurality of guide rails that guide forward movement of the brush and are positioned at both end sides of the roller. The method of claim 2, wherein the brush, The distance between the end of the brush and the neighboring brush is set to be longer than the average length or the average length of the grain, Particle quality discrimination system, characterized in that arranged next to the neighboring brush, or the end forming shape with the surrounding brush in a lattice manner. The method of claim 2, wherein the particulate quality determination system, And a transfer control unit for controlling driving of the motor according to an input of a management personnel while supplying driving power to the sample transfer / separation unit. The method of claim 1, wherein the light source unit, A first light source unit for irradiating light onto the sample in the same direction as the light source detection unit; And a second light source unit that faces the light source unit and the sample therebetween, and which directs light in the direction of the sample. The method of claim 6, wherein the light detector, And a quantity of reflected light irradiated from the first light source and reflected on the sample, and an amount of transmitted light irradiated from the second light source and transmitted through the sample. The method of claim 1, wherein the management system, A power supply unit for supplying driving power to the light source unit and the sample transfer / separation unit; An input unit for receiving input information from an administrative agent, A display unit for outputting a display screen; A main control unit controlling driving of the light source unit and the sample transfer / separation unit according to the input information input through the input unit; And an analyzer configured to acquire an image of each grain of the sample based on the amount of reflected light and transmitted light received from the light detector, and to calculate and process the sample based on the image. The method of claim 8, wherein the analysis unit, Converts the transmitted light received from the light detector into a gray level, applies an anisotropic diffusion filter, converts the gray level image into a binarized image, and then converts the binary inverse, and numbers each of the grains. A grade quality discrimination system characterized by extracting shape information and discriminating the quality on the basis of the quality judgment criteria. The method of claim 9, wherein the analysis unit, The Harris corner detection is used to distinguish each grain based on two points where the corner points are closest to each other in the binary inverted image, and to clean up the edges by arranging the blurry values generated at the edges of the grain. Particle grade discrimination system characterized by. The method of claim 9, wherein the analysis unit, The image of each grain is stored based on the transmitted light and the reflected light image of each grain of the sample, the number information and each grain boundary value assigned to each grain, and the position information of each grain to be extracted. And calculating the circularity and calculating the circularity, and then calculating the circularity and extracting the shape information. The method of claim 11, wherein the analysis unit, The circularity is calculated using Equation 1 below, and the closer the grain is to the circle, the closer to the value of the particulate matter discrimination system, characterized in that 1.0. [Equation 1]

According to claim 1, The light source unit, Irradiating medium or near infrared with the sample, The light detector The fine grain quality discrimination system which detects the wavelength of the said medium infrared rays or near infrared rays which permeate | transmits the said sample. The method of claim 13, wherein the management system, And a qualitative and quantitative analysis system for analyzing qualitative and quantitatively by inspecting atomic spectra based on the wavelength of the medium or near infrared rays detected by the light detector. In the quantitative analysis system of rice grains using spectroscopic analysis, A light source unit for irradiating medium or near infrared light with a sample; A light detector for detecting the medium or near infrared rays irradiated from the light source unit and passing through the sample; Quantitative analysis system of rice flour using a spectroscopic method including a qualitative analysis based on the middle infrared rays detected by the light detector, a quantitative analysis based on near infrared rays, and correcting / modeling qualitative and quantitative values. In a grain grade discrimination method, Injecting a sample into the mirim grade determination system; Irradiating light to the injected sample; Detecting transmitted light and reflected light that pass through the sample; Acquiring shape information of an image based on the transmitted light and the reflected light; And determining the quality of the sample based on the image. The method of claim 16, wherein the shape information of the image comprises: An image acquiring step of acquiring an image based on the amount of reflected light and the amount of transmitted light; A pretreatment step of converting the transmitted light into a gray level and applying an anisotropic diffusion filter; A region dividing step of converting the gray level image into a binarized image; A binary inverse transformation step of substituting a background and a grain divided in the binarized image; A labeling step of numbering each grain; And a shape information extracting step of extracting shape information of each grain. The method of claim 17, Edge splitting step using Harris corner detection to distinguish each grain based on two points where the corner points are closest to each other; The method of claim 1, further comprising a border arranging step for arranging a blurry value generated in the image border of each grain. The method of claim 17, wherein the extracting shape information comprises: Receiving the transmitted light and the reflected light image for each grain, the number information assigned to each grain, and each grain boundary value; Extracting location information for each number of each grain; Storing as a single image based on the small and last positions of each grain; Calculating the center point and the length of each grain; Calculating the area of each grain in pixel units; Calculating a perimeter length based on the length of the outline of each grain; Comprising a step of calculating the circularity of each grain. The method of claim 16, wherein irradiating the light, The first light is irradiated in the direction of the same sample as the light detector for detecting the light to generate the reflected light, and the second light is generated by facing the light detector and the sample to generate the transmitted light. How to determine grain grades. The method of claim 16, Irradiating the sample material with medium or near infrared rays; Detecting the medium or near infrared rays passing through the sample; Qualitative analysis based on the infrared light; Quantitative analysis based on the near infrared ray, Comprising the step of correcting / modeling the numerical value of the qualitative and quantitative analysis results.

Images