KR20110018562A - Cow carcass grade decision apparatus and method - Google Patents

Abstract

PURPOSE: A small conductor meat quality grade determining unit and a method thereof using through an image process are provided to increase the reliability about a rating result and contribute to farmhouse income increase according to an objective apparatus. CONSTITUTION: A small conductor picking up device(110) picks up the small conductor. An image data input part(121) inputs the image data from the small conductor. A color correction module(122) corrects a color correction regression equation. An image process module(123) extracts interest region data of the fixed size. A rating module determines the meat quality grade of the small conductor by comparing the stored standard data with the interest region data.

Description

Small carcass quality grading apparatus and method {COW CARCASS GRADE DECISION APPARATUS AND METHOD}

The present invention relates to an apparatus and method for determining the meat quality of small conductors, and more particularly, to an apparatus and method for determining the quality of small body meat using image processing.

In general, domestic beef grades are determined by classifying them into meat grades and meat grades. Meat grade is to determine meat quality as 1 ++, 1+, 1, 2, 3 grades according to muscle fatness, meat color, fat color, texture, and maturity, and meat grade is weight, backfat, etc. The thickness and the cross-sectional area of the fillet are combined to determine grades A, B, and C.

Beef meat is a protein food and its nutritional value is also important, but beef is the most important food taste and softness. Intramuscular fat was the most important aspect of the taste of beef, and maturity (age) and muscle texture and texture were related to the age related to beef taste. In order to enhance visual satisfaction, flesh color requires indigenous crimson and local color requires white. Meat color is an indicator of meat quality disqualification by abnormal meat. The intramuscular fat map was determined as a preliminary grade by comparing the degree of fat distribution shown in the section of the longest muscle muscle on the grading panel with the criteria of FIG. 19.

Meat color is determined on the grading scale by comparing the meat color of the cut root muscle section with the meat color standard according to FIG. 20. The fat color is determined by comparing the color of the intramuscular fat, the surrounding muscle fat and the back fat of the dorsum muscle cross section with the fat color standard shown in FIG.

The meat grade is calculated by the meat index according to the backfat thickness, the closest root area, and the carcass weight. If the index is over 67.5, the meat grade is classified into A grade, 62.0 ~ 67.5 B grade, and C grade less than 62. The meat grade judging area is determined by incision between the last spine and the first lobe of the left semiconductor, which is divided into two after the slaughter of cattle.

Intramuscular fat is one of the most important factors in determining the grade of meat. Intramuscular fat is known to affect the year, juiciness, and flavor.

At present, the intramuscular fat standard of cattle or pigs in Korea is different in crude fat content by 1-2%, but the crude fat content of ± 2% in muscle at the standard area is virtually indistinguishable due to visual recognition limits. In addition, the small carcasses have different characteristics of muscle fiber and visual expression of intramuscular fat according to breed, gender, and age. However, the standard lookup table has a single injection expression shape, which may cause errors in visual interpretation among the measurers.

Therefore, if the objective device judgment method is developed and distributed rather than relying on the subjective visual judgment, it is expected to contribute to the increase of farm income, such as promoting the production of high quality meat and differentiating from imported meat by increasing the reliability of the grading result.

The present invention has been proposed to satisfy the above needs, and an object of the present invention is to automate grading by identifying intramuscular fat, flesh color, and fat color using image processing after imaging small conductors with a camera. It is to provide an apparatus and method for judging the quality of small carcasses using image processing.

In order to achieve the above object, the apparatus of the present invention comprises a small conductor imaging unit for imaging the small conductor; An image data input unit configured to receive image data from the small conductor imaging unit and store the image data; A color correction module that initially receives image data of the color plate from the image data input unit, receives theoretical values of the color plate, and calculates a color correction regression equation by regression analysis; An image processing module for correcting color of the image data input through the image data input unit according to the color correction regression equation and extracting ROI data having a predetermined size; And initially receiving the ROI data of the image of each grade of the graded standard sample from the image processing module, generating and storing standard data for determination, and during measurement, a small conductor image to be determined through the image processing module. It is characterized in that it comprises a grading module for receiving a region of interest of the data to compare with the stored standard data to determine the meat grade of the small conductor.

The small conductor quality grading apparatus further includes an output unit for outputting a determination result of the grading determination module, an input unit for inputting an operation, and the small conductor image pickup unit includes an illumination LED for illuminating the small conductor and the illumination. A camera for capturing an image of the small conductor illuminated by the LED, and a dome-shaped shading means for blocking external illumination and maintaining uniform illuminance on the surface of the small conductor.

In addition, in order to achieve the above object, the method of the present invention comprises a first step of generating standard data for meat grade determination of small carcase; A second step of receiving an image of the small conductor to be determined; A third step of performing image processing by transforming the input RGB image of the small conductor into a Lab image; Extracting region of interest data from the processed image; A fifth step of determining a meat grade by comparing the ROI data with standard data; And a sixth step of outputting the determined meat grade.

According to the present invention, in determining the meat grade of small carcasses, it can be judged according to an objective device rather than relying on the subjective visual judgment, thereby increasing the reliability of the grading result and increasing farm income, such as promoting production of high-quality meat and differentiating from imported meat. It is possible to make a significant contribution to this, and to achieve the early practical use of small scale quality assessment technology, which is still in the research stage in other countries, can provide an opportunity to lead the field.

The technical problems achieved by the present invention and the practice of the present invention will be more clearly understood by the preferred embodiments of the present invention described below. The following examples are merely illustrative of the present invention and are not intended to limit the scope of the present invention.

1 is a block diagram showing the configuration of a small conductor quality rating apparatus according to the present invention, Figure 2 is a detailed view of the small conductor imaging unit shown in FIG.

As shown in FIG. 1, the small conductor quality grade determining apparatus 100 according to the present invention includes a small conductor imaging unit 110 for imaging a small conductor (tenderloin), an image data input module 121, and color correction. Module 122, an image processing module 123, a rating determination module 124, an input unit 130, an output unit 140, the image data input module 121, the color correction module 122, the image destination The remodule 123 and the rating determination module 124 may be implemented as a program driven on the main computer 120. Typically, the main computer 120 is composed of a MCU and peripheral circuits, memory, operating system, application program.

Referring to FIG. 1, as illustrated in FIG. 2, the small conductor imaging unit 110 includes an illumination LED 113 for illuminating the small conductor 10 and a small conductor (illuminated by the illumination LED 113). A camera 114 for capturing the image of 10) and a dome-shaped light blocking mechanism 111 for blocking external illumination and maintaining uniform illuminance on the surface of the small conductor. The camera 114 positioned on the light blocking device 111 is installed in the body 112, and the handle 112 is attached to the body 112 so that the user grasps the handle 115 and moves the light blocking device 111. It is possible to image the desired part of 10). Although not shown in detail in the drawing, the optical conductor (reflective film, lens, etc.) for improving image quality may be added to the small-conductor imaging unit 110, and may be connected to the main computer 120 through a USB cable. The illumination LED 110 uses a yellow illumination LED, and in some cases, may use a light source having a wavelength of approximately 680 to 720 nm (preferably 694 nm) having a high degree of separation between lean meat and fat. As the camera 114, a CMOS camera, a CCD camera, or the like may be used, and the small conductor imaging unit 110 may be wirelessly connected to the main computer 120 such as Bluetooth or Zigbee.

The image data input module 121 receives image data from the small conductor imaging unit 110 and stores the image data. The color correction module 122 initially captures a color plate known as RGB to the small conductor imaging unit 110. The image data of the color plate is input through the image data input module 121 and the theoretical value of the color plate is input through the input unit 130 to calculate the color correction regression equation by regression analysis. When the color correction regression equation is obtained in advance and stored in the memory during the device manufacturing step, the color correction module 122 may be omitted or may be operated only during the calibration step.

The image processing module 123 performs color correction on the image data input through the image data input module 121 according to the color correction regression equation, removes noise, and coordinates the input RGB (Red Green Blue) image to Lab image. Extracts ROI data of a set size. The size of the ROI may be 60 mm or 40 mm in diameter and set by the user in the setting step. And Lab image is to minimize the difference in color according to the output device. a and b contain information about Hue, and the L axis contains all information about Luminosity.

The grading determination module 124 receives the ROI data of each grading image of the grading standard sample from the image processing module 123, generates and stores standard data for determination, and measures the image processing module. In step 123, the region of interest of the small conductor image to be determined is input and compared with the stored standard data to determine a meat grade of the small conductor. At this time, the standard data for grading is composed of the myocardial fat standard data for each grade, the meat color standard data, and the fat color standard data. The standard data for meat grade determination may be obtained in advance and stored in a memory at the manufacturing stage. In addition, the grading determination module 124 compares the ROI data with the intramuscular fat standard data for each grade to determine the musculoskeletal fat, compares the ROI data with the meat color standard data for each grade, and determines the flesh color grade. The fat color grade is determined by comparison with the fat color standard data for each grade. Finally, the muscle quality is determined to be 1 ++, 1+, 1, 2, and 3 grades by integrating intramuscular fat, meat color, and fat color determination results.

The output unit 140 outputs the determination result of the rating determination module 124, and the input unit 130 inputs data or manipulations such as color plate theoretical values or ROI sizes. The output unit 140 may be implemented by any one or more of an LCD, a label printer, or a communication means for transmitting the determination result to the outside, and the input unit 130 may be implemented by a keypad or a touch screen.

In addition, the small conductor quality rating apparatus 100 according to the present invention includes an image data input module 121, a color correction module 122, an image processing module 123, a rating determination module 124, an output unit 140, The input unit 130 forms a main body, and the small conductor imaging unit 110 may be manufactured to be movable, so that the main body and the small conductor imaging unit 110 may be connected to a wired / wireless communication line to transmit an image signal. In addition, one small portable imaging unit 110, the image data input module 121, the color correction module 122, the image processing module 123, the grading determination module 124, the output unit 140 is portable It can also be produced in the form of a terminal.

3 is a flowchart illustrating a process for determining the quality of the small body meat according to the present invention, FIG. 4 is a detailed flowchart of the color correction procedure shown in FIG. 3, and FIG. 5 is a view of the card for color correction used in the color correction of FIG. 4. Yes.

Referring to FIG. 3, in the initial setting step S1, necessary information such as the size of the ROI is input. In the color correction regression calculation step S2, when the color correction regression equation is not obtained, the color correction regression equation is calculated according to the procedure shown in FIG. 4. Since the image color of the small conductor may vary depending on the lighting conditions, color correction is essential for accurate determination. Referring to FIG. 4, 24 color plate images are obtained by capturing a standard color plate of which RGB values are known as shown in FIG. 5 through the small-conductor imaging unit 110 (S21). Subsequently, RGB of the color plate is extracted from the color plate image, and the RGB theoretical values of the color plate obtained by the standard color judging device are received through the input unit 130 (S22 and S23). In addition, the RGB value of the extracted color plate and the theoretical value are multi-regressively analyzed to calculate and store the color correction regression (S24, s26). This color correction regression equation is then used to correct the color of the image input from the image processing module 123.

Referring again to FIG. 3, when there is no standard data for determining the grade, the intramuscular fat degree, meat color, and fat color standard data for each grade are obtained according to the procedure shown in FIG. 6 (S3).

Referring to FIG. 6, first, a sample image for each grade is obtained, image data is input, color correction is performed according to a color correction regression equation, noise is removed, and the RGB image is coordinate converted to a Lab image (S31 to S35). After extracting the region of interest, the standard intramuscular fat standard data for each grade, the meat color standard data for each grade, and the fat color standard data for each grade are calculated (S36 ~ S39).

7 is a sample image for each grade obtained by the size of the ROI 60mm in accordance with the present invention, Figure 8 is a measurement of intramuscular fat obtained by 60mm of the size of the ROI in accordance with the present invention, Figure 9 is the present invention In accordance with the present invention, the measured color is obtained by using the size of the region of interest as 60 mm, and FIG.

When the diameter of the region of interest is 60mm, the intramuscular fat level was measured for each of the samples as shown in FIG. 7 from grade 1 ++ to grade 3 using the apparatus of the present invention. When the ROI was 60 mm in diameter, it was possible to distinguish 1 ++ grades, but the remaining grades contained many fat bands, which made it difficult to classify them. The total number of samples used was 203, which was slightly different for each grade. For the low grade 3, 14 samples were measured twice. In this example, the 1 ++ grade showed a marked difference compared to the other grades, but for the other grades, it was not easy to distinguish the intramuscular fat. The reason is that the 1 ++ grade contains a lot of intramuscular fat, unlike other meat grades, but the 1+ grade and the first grade were determined to change the amount of intramuscular fat significantly depending on the presence or absence of a fat band. Therefore, in determining the meat grade, it is judged that the remainder of the fat strip removed must be used. Here, the distribution of the meat color and the fat color for each grade is as shown in FIGS. 9 and 10. As the grade was higher, there was no tendency to increase a value or b value in the distribution of flesh color, and there was no correlation in distribution of fat color. Meat color and fat color set the standard for grade down adjustment rather than grade difference. That is, if it falls above or below a certain value, the grade is adjusted downward.

FIG. 11 is a sample image for each grade obtained by using a size of an ROI of 40 mm according to the present invention, FIG. 12 is a measurement of intramuscular fat obtained by using a size of an ROI of 40 mm according to the present invention, and FIG. In accordance with the present invention, the measured color is obtained by using the size of the ROI as 40 mm, and FIG. 14 shows the measured measured fat color by using the size of the ROI as 40 mm.

In the sample as shown in FIG. 11, when the ROI was 40 mm in diameter, only the meat part was measured to avoid the fat band, and the intramuscular fat distribution of each grade was as shown in FIG. 12. Although it was difficult to distinguish between grades 2 and 3, it was possible to distinguish between grades 1 ++, 1+, and grade 1.

Each grade is not clearly distinguished, although the area of interest is small. It is difficult to expect linearity. However, in the present invention, it was confirmed that intramuscular fat can be classified by grades by eliminating the fat band by reducing the region of interest. Measurements are possible, and on the other hand, better results can be obtained by using the 694 nm wavelength band shown in the measurement results of intramuscular fat using hyperspectral image processing. Here, the measurement results of meat color and fat color for each grade are as shown in FIGS. 13 and 14. Since most of the regions do not contain a fat band, the a and b values of meat color seem to have different characteristics, and the distribution of fat color also tends to be different. Therefore, using this data, it is possible to determine the objective color grade in addition to the existing grade chart.

Referring back to FIG. 3, after the standard data for grading determination is obtained, the small conductor 10 on which the user actually determines the meat grade is placed on a table and the small conductor 10 is picked up by the small conductor imaging unit 110. (S4, S5). In this case, in order to improve the accuracy of the determination, it is preferable to image the part without fat band as much as possible, and attach the measurement button for imaging to the small-conductor image pickup unit 110 and press the measurement button 116 to capture the small-conductor image. It is preferable.

After receiving the image data of the small conductor to be determined through the small conductor imaging unit 110, processing the image according to the procedure illustrated in FIG. 15, and determining the meat quality of the small conductor according to the procedure illustrated in FIG. 16. (S6 to S8).

15 is a detailed flowchart illustrating a procedure of processing a small conductor image to be determined according to the present invention.

Referring to FIG. 15, colors are corrected by three color correction regression equations for R, G, and B in the color correction step S71, and noise is removed through 3 × 3 mean filtering in the noise removal step S72. do. In the color coordinate conversion step (S73), the RGB image is converted into a Lab image, and then the region of interest is extracted. In the region of interest extraction step (S74), only the image of the center portion is extracted except for the outside of which there are few light sources.

16 is a detailed flowchart illustrating a procedure for determining the meat grade of small carcasses according to the present invention.

Referring to FIG. 16, after extracting fat at a specific value of Lab in a region of interest, the area of fat and non-fat region in the region of interest is calculated, and the distribution of fat and non-fat regions is calculated to determine intramuscular fat. (S81, S82). Then, meat color is determined, and fat color is determined (S83, S84). That is, the flesh color grade is determined by comparing the ROI data with the meat color standard data for each grade, and the fat color grade is determined by comparing the ROI data with the fat color standard data for each grade.

Figure 17 is a graph showing the standard deviation of the intramuscular fat by grade according to the present invention, Figure 18 is an example of the spectrum of fat and lean meat when the wavelength is different in the present invention.

Intramuscular fat when the fat band is included in the region of interest can be measured by removing the large fat band using image processing and then measuring the intramuscular fat for the remaining parts, but the fat band is often associated with normal intramuscular fat. The result of measuring the intramuscular fat after avoiding the part with is as shown in FIG. Referring to FIG. 17, it can be seen that the difference between grades becomes clearer as the fat strip is removed. Therefore, it is preferable to remove the fat strip when using image processing.

18 shows the average reflection spectra of fats and lean meats obtained by varying the wavelength in one graph. As a result of comparing the spectra of FIG. 18, it can be seen that there is almost no difference in reflectance between fat and lean meat in the wavelength band of 467 nm or less, and the largest spectral difference between fat and lean meat in the wavelength band of 694 nm. Therefore, a light source having a wavelength of approximately 680 to 720 nm (preferably 694 nm) having a high degree of separation between lean meat and fat may be used in the small conductor imaging unit 110.

The present invention has been described above with reference to one embodiment shown in the drawings, but those skilled in the art will understand that various modifications and equivalent other embodiments are possible therefrom.

1 is a block diagram showing the configuration of a small carcass quality grading apparatus according to the present invention,

FIG. 2 is a detailed view of the small conductor imaging unit shown in FIG. 1;

3 is a flow chart showing a small carcass quality grading procedure according to the present invention,

4 is a detailed flowchart of a color correction procedure shown in FIG. 3;

5 is an example of a color correction card used in the color correction of FIG.

6 is a flowchart illustrating a procedure of acquiring standard data for each grade according to the present invention;

7 is a sample image for each grade obtained by the size of the region of interest according to the invention 60mm,

8 is a measurement of intramuscular fat obtained by making the size of the ROI 60 mm according to the present invention;

9 is a color measured value obtained by making the size of the ROI 60 mm according to the present invention;

10 is a measurement of fat color obtained by making the size of the ROI 60 mm according to the present invention;

11 is a sample image for each grade obtained by the size of the ROI 40mm according to the present invention,

12 is a measurement of intramuscular fat obtained by making the size of the ROI 40 mm according to the present invention;

13 is a color measured value obtained by setting the size of the ROI to 40 mm according to the present invention;

14 is a measurement of fat color obtained by using a size of a region of interest according to the present invention as 40 mm,

15 is a detailed flowchart illustrating a procedure of processing a small conductor image to be determined according to the present invention;

16 is a detailed flowchart illustrating a procedure for determining meat grade of small carcasses according to the present invention;

17 is a graph showing the standard deviation of intramuscular fat by grade according to the present invention;

18 is an example spectrum of fat and lean meat when the wavelength is different in the present invention,

19 is based on the cattle carcass fat grading according to the livestock product grading detailed criteria,

20 is a small carcass color standards according to livestock product grading detailed criteria,

Figure 21 is a small carcass fat color according to livestock grading detailed criteria.

DESCRIPTION OF THE REFERENCE NUMERALS

10: small conductor 100: small body meat grade determination device

110: small conductor imaging unit 121: image data input module

122: color correction module 123: image processing module

124: rating determination module 130: input unit

140: output unit

Claims (10)

A small conductor imaging unit for imaging the small conductor; An image data input unit configured to receive image data from the small conductor imaging unit and store the image data; A color correction module that initially receives image data of the color plate from the image data input unit, receives theoretical values of the color plate, and calculates a color correction regression equation by regression analysis; An image processing module for correcting color of the image data input through the image data input unit according to the color correction regression equation and extracting ROI data having a predetermined size; And Initially, from the image processing module, the ROI data of each grade of the graded standard sample is input to generate and store standard data for determination. And a grading module that receives a region of interest data and compares the stored standard data with a grading grade of the small carcase. According to claim 1, wherein the small carcass quality grading apparatus And an output unit for outputting a determination result of the rating module and an input unit for inputting an operation. And the output unit is any one of an LCD, a label printer, and communication means for transmitting the determination result to the outside, and the input unit is a keypad or a touch screen. The small conductor imaging unit according to claim 1 or 2, wherein An illumination LED for illuminating the small conductor, a camera for capturing an image of the small conductor illuminated by the illumination LED, and a dome-shaped shading means for blocking external illumination and maintaining uniform illuminance on the surface of the small conductor Small body meat grade determination device, characterized in that consisting of. According to claim 2, wherein the small carcass quality grading apparatus The image data input unit, the color correction module, the image processing module, the rating determination module, the input unit, and the output unit form a main body, and the small conductor imaging unit is manufactured to be movable, so that the main body and the small conductor imaging unit are wired / wireless communication. Small body quality class determination device characterized in that connected to the line to transmit the video signal. According to claim 2, wherein the small carcass quality grading apparatus Small conductor image quality grading apparatus characterized in that the small conductor imaging unit, the image data input unit, the color correction module, the image processing module, the rating determination module, the input unit, and the output unit is manufactured in the form of a single portable terminal. The image processing module of claim 1, wherein the image processing module is An apparatus for measuring the quality of a small body material, characterized in that for converting an input RGB image into a Lab image. According to claim 1 or claim 2, wherein the small carcass quality grading apparatus is Obtaining the standard data for the color correction regression and meat grade determination in advance, the small carcass meat grade determination device, characterized in that provided in the manufacturing step stored in the memory. A first step of generating standard data for meat grade determination of the small carcase; A second step of receiving an image of the small conductor to be determined; A third step of performing image processing by transforming the input RGB image of the small conductor into a Lab image; Extracting region of interest data from the processed image; A fifth step of determining a meat grade by comparing the ROI data with standard data; And And a sixth step of outputting the determined meat grade. The method of claim 8, wherein the small carcass quality grading method is And a color correction step for correcting the color of the image input in the second step. The method of claim 8, wherein the standard data is composed of the myopia fat standard data for each grade, meat color standard data, fat color standard data, In the fifth step, the intramuscular fat is determined by comparing the ROI data with the intramuscular fat standard data for each grade, and the ROI data is determined by comparing the ROI data with the meat color standard data for each grade, and the ROI data is determined with the fat color standard for each grade. A small carcass quality grading method according to claim 1, wherein the grading of fat color is compared with data.

Images