Detection and analysis method of handheld white feather broiler quality detection equipment
Through the image processing method of handheld white-feathered broiler quality detection equipment, the problem of time-consuming and inaccurate manual inspection is solved, and efficient and accurate judgment of chicken quality is achieved.
Patent Information
- Application Number
- CN202510154170.3
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2025-02-12
- Publication Date
- 2025-07-04
- Estimated Expiration
- Not applicable · inactive patent
AI Technical Summary
In the prior art, manual detection of chicken quality takes a lot of time and is inaccurate. Equipment detection is easy to change the chicken environment, and it is difficult to accurately distinguish parts of the chicken with similar colors such as yellow skin, blood stasis and hair roots.
The handheld white-feathered broiler quality detection equipment is used to take chicken photos and perform grayscale and binarization processing, and weighted average is performed with chicken color characteristics. The number of pixels in the connected area is extracted by image processing technology, color comparison and analysis is performed, and quality is judged based on the database.
It improves the efficiency and accuracy of chicken quality detection, reduces misjudgment and misjudgment, and avoids the impact of manual intervention on chicken quality.
Smart Images
Figure CN120253710A_ABST
Abstract
Description
Technical Field
[0001] The present invention belongs to the technical field of food safety detection, and particularly relates to a detection and analysis method for a handheld quality detection device for white - feather broilers. Background Art
[0002] The state and enterprises have clear standards for the appearance grading of chicken cut - up meat. The main criteria for judgment are to detect yellow skin, congestion, and visible hair roots with a diameter less than 2 mm. The following are the commonly used detection standards in chicken production enterprises: Grade - one chicken cut - up meat: Regardless of the part, there shall be no yellow skin, congestion, or visible hair roots with a diameter less than 2 mm.
[0003] Grade - two chicken cut - up meat: The yellow skin does not exceed 4 places per 10 kg, and the total area ≤ 4 cm2. The congestion area ≤ 1 cm2, and does not exceed 2 places per 10 kg. For visible hair roots with a diameter less than 2 mm, there are ≤ 2 roots per whole leg, ≤ 1 root per lower leg, ≤ 2 roots per whole wing, ≤ 1 root per wing root and wing middle part, and ≤ 1 root per breast piece.
[0004] Grade - three chicken cut - up meat: The yellow skin does not exceed 10 places per 10 kg, and the total area ≤ 10 cm2. The congestion area ≤ 1 cm2, and does not exceed 10 places per 10 kg. For visible hair roots with a diameter less than 2 mm, there are ≤ 3 roots per whole leg, ≤ 2 roots per lower leg, ≤ 3 roots per whole wing, ≤ 2 roots per wing root and wing middle part, and ≤ 2 roots per breast piece. Through the grading of chicken cut - up meat, problem - chicken can be preliminarily screened out, improving the safety of people's food.
[0005] Due to the special appearance structure of chicken cut - up meat, various problems often occur during meat quality detection. For example, yellow skin and congestion are sometimes similar in color to chicken meat, making it difficult to distinguish, resulting in problems such as unclear judgment and inability to distinguish when using existing conventional machine vision detection equipment for detection. When using manual identification, it takes a lot of time, and the manual factor also has a great influence on the quality judgment; also, because the color of hair roots is close to transparent, it is easy to be ignored during machine detection or manual identification.
[0006] Through the above - mentioned background analysis, it is found that the following problems exist in the existing technologies: Problem one: Manual grading is adopted. By visually identifying yellow skin, congestion, and hair roots, and using a caliper to measure and manually calculate and statistically analyze data for grading, it takes a lot of time.
[0007] Problem two: The intervention of humans will have an impact on the chicken meat itself, and the influence of environmental factors is likely to cause changes in the quality of chicken meat.
[0008] Problem three: For parts such as congestion that are similar in color to the chicken meat itself, simply observing with the naked eye is likely to be overlooked, affecting the judgment of the quality of the chicken meat itself and reducing the accuracy of the results. Summary of the Invention
[0009] In view of this, the present invention provides a detection and analysis method for a handheld white - feather broiler quality detection device, which can solve the problems that the existing manual detection takes a lot of time and cannot guarantee the accuracy of detection; at the same time, it can solve the problem that the current device needs to change the environment of the chicken to be detected when detecting chicken, and the detection means has hysteresis.
[0010] The present invention is implemented as follows: The present invention provides a detection and analysis method for a handheld white - feather broiler quality detection device, specifically including: S1: Collect photos of the chicken to be detected, store and upload the photos to the detection terminal; S2: Image pre - processing. Gray - scale the image obtained by front - end acquisition in step S1, and perform weighted averaging on the red, green, and blue components according to the color characteristics of the chicken to obtain a more reasonable gray - scale image, which is convenient for subsequent comparison and matching with the database; S3: Binarize the pre - processed image in step S2 to reduce the difficulty of comparison; S4: Connect the pixels in the region of the binarized image in step S3 according to the chicken quality evaluation index, and obtain the total value of the pixel numbers in the connected region under each index; S5: Synchronously perform color comparison and analysis on the regions with large color value differences in the evaluation index; S6: Summarize the data of the above steps, count the pixel data of each index region, and calculate the actual area of each index region; S7: Compare the area data obtained in step S6 with the preset data in the database to judge the chicken quality, and send the analysis result to the front - end.
[0011] Based on the above technical solution, a handheld white - feather broiler quality detection device of the present invention can also be improved as follows: Among them, in step S4, the specific steps to obtain the pixel number of the evaluation index region image are as follows: According to the gray - scale reference threshold of the white - feather broiler feather root, the gray - scale reference threshold of congestion, and the gray - scale reference threshold of yellow skin preset in the background, further image processing is performed on the image data after binarization in step S3. The specific steps include: S41: Scan the image and select the current pixel point B(x, y)=1; S42: Take B(x, y) as the initial position of the pixel point, assign a label to this pixel point, and then push all adjacent foreground pixels of this pixel point onto the stack; S43: Pop the top - of - stack pixel, assign it the same label, and then push all adjacent foreground pixels of this top - of - stack pixel onto the stack; Repeat step S43 until the stack is empty; S45 Output the results to obtain the hair root gray-scale connected region image, the congestion gray-scale connected region image, and the yellow skin gray-scale connected region image respectively, and calculate the number of pixels within the connected regions.
[0012] Furthermore, in step S5, the specific steps of color comparison are as follows: S51 Separate the congestion gray-scale connected region image and the yellow skin gray-scale connected region image obtained in step S45, and respectively obtain the central pixel coordinate values (X1, Y1), (X2, Y2),..., (Xn, Yn) of each connected region; S52 Extract the 8 pixel points adjacent to the coordinate values with the above coordinate values as the center, and calculate the color difference of the 9 pixel points including the 8 pixel points and the central pixel point; S53 Quantify the color information of the pixel region obtained in S52, and generate a color feature vector based on the percentages of the total pixels of various colors of red, green, and blue. A percentage greater than 90% is considered to meet the determination standard.
[0013] The present invention also provides a handheld white - feather broiler quality detection device, including a detector housing, a display screen, a shooting lens, a switch button, and a WiFi module; the detector housing is divided into a handle and a main body; a battery compartment is rotatably installed below the handle; the switch button is fixedly installed on the handle; the display screen, the shooting lens, and the WiFi module are fixedly installed on the main body; the switch button is electrically connected to the display screen, the shooting lens, and the WiFi module; Among them, this device uses the detection and analysis method of the above - mentioned handheld white - feather broiler quality detection device to detect the quality of white - feather chicken meat.
[0014] The beneficial effects of adopting the above - mentioned improved scheme are: by extracting the center point and combining the 9 adjacent pixel points through the center point coordinates, and comparing with the corresponding indicators in the original database, compared with directly comparing the basic image with the corresponding indicators in the original database, it can greatly reduce the calculation amount while avoiding the loss of credibility and accuracy of the results.
[0015] Compared with the prior art, the beneficial effects of the detection and analysis method of the handheld white - feather broiler quality detection device provided by the present invention are: the present invention uses a portable handheld shooting device to take pictures of the chicken cut meat, upload it to the cloud platform, and through the background artificial intelligence vision algorithm, statistically calculate the areas of the yellow skin and congestion and the number of hair roots, and grade through the threshold standard, which can greatly improve the work efficiency and accuracy, and reduce the misjudgment and missed judgment rates. Brief Description of the Drawings
[0016] Figure 1 It is the overall flowchart; Figure 2 It is a flowchart of method steps; Figure 3 It is a schematic diagram of the overall structure.
[0017] In the accompanying drawings, the list of components represented by each reference numeral is as follows: 1. Detector housing; 2. Display screen; 3. Shooting lens; 4. Switch button; 11. Handle; 12. Main body; 13. Battery compartment. Detailed implementation manners
[0018] In order to make the objectives, technical solutions and advantages of the present invention clearer and more understandable, the present invention will be further described in detail below with reference to the accompanying drawings. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not used to limit the present invention.
[0019] As Figure 1 , shown in FIG. 2, it is a detection method of a handheld white - feather broiler quality detection device provided by the present invention. This method specifically includes: S1. Collect photos of the chicken to be detected, store and upload the photos to the detection terminal; S2. Image pre - processing: perform grayscale processing on the image collected at the front - end in step S1, and perform weighted averaging on the three RGB components according to the color characteristics of the chicken to obtain a more reasonable grayscale image, which is convenient for subsequent comparison and matching with the database; S3. Perform binarization processing on the pre - processed image in step S2 to reduce the difficulty of comparison; S4. Connect the pixels of the binarized image in step S3 according to the chicken quality evaluation index, and obtain the total pixel number of the connected regions under each index; S5. Simultaneously perform color comparison and analysis on the regions with large color value differences in the evaluation index; S6. Summarize the data of the above steps, count the pixel data of each index region, and calculate the actual area of each index region; S7. Compare the area data obtained in step S6 with the preset data in the database, judge the chicken quality, and send the analysis result to the front - end.
[0020] Among them, in the above solution, the specific steps of S1 for collecting pictures include: S11. Prepare the chicken to be detected and place it flat in the 400*300mm detection area; S12. Hold the detection device by hand, make the detection lens face the chicken, and take a picture of the chicken; S13. Press the shooting button, and the camera takes a picture and uploads it to the terminal.
[0021] Further, in the above solution, according to the gray-scale reference threshold of white feather chicken root preset in the background, the image after binarization processing in S3 is extracted. The specific steps for obtaining the number of pixels in the gray-scale area image of white feather chicken root are as follows: S41 Scan the picture taken and uploaded in S1, select a pixel point A(x, y) as the initial pixel point, assign a label to this pixel point, and use this as the initial pixel position. Press all the foreground pixels adjacent to this initial pixel position into the stack in a clockwise manner; S42 Pop the top pixel of the stack, assign the same label to this pixel, and in the same way as in S41, press the foreground pixels adjacent to the top pixel of the stack into the stack again; S43 Repeat step S42. Finally, there are no more foreground pixels around the pixel point, and the stack starts to repeatedly pop the pixels at the top position of the stack outwards until the stack is empty; S44 In this way, a continuous image of the gray-scale area of white feather chicken root with the same label is obtained, calculate the number of labels in this image, and obtain the number of pixels Number1 in the gray-scale area of white feather chicken root.
[0022] Further, in the above solution, the number of pixels Number2 in the yellow skin area and the number of pixels Number3 in the congestion area are obtained in the same way.
[0023] Further, in the above solution, the specific steps for color comparison and analysis by color value in step S5 are as follows: S51 Obtain the continuous images of the above yellow skin area and the congestion area, and obtain the central pixel coordinates (X1, Y1), (X2, Y2) …… (Xn, Yn) of each continuous image; S52 According to the central coordinates obtained in S51, extract the 8 pixel points around this central coordinate, and combine these 8 pixel points and the central pixel point to form a total of 9 pixel points to splice new pixel areas B1, B2 …… Bn; S53 Quantify the color information of the above pixel areas B1, B2 …… Bn, and generate a color feature vector based on the red, green, blue colors and the percentage of the corresponding pixels in the total pixels.
[0024] Further, in the above solution, the specific steps for calculating the area in S6 are as follows: According to the inspection area of 400 * 300 mm and the camera resolution of 5472 * 4648, calculate the mechanical area of each pixel point; According to the number of pixels Number1, Number2, Number3 in the connected areas of the chicken root, yellow skin, and congestion, the actual area S of each area can be calculated as S = Number * (400 * 300) / (5472 * 4648); Calculate the actual areas S1, S2... Sn of the continuous images in each region according to the above formula.
[0025] Such as Figure 3 , which is a handheld white - feather broiler quality detection device provided by the present invention. Specifically, the device includes a detector housing 1, a display screen 2, a shooting lens 3, a switch button 4, and a WiFi module; the detector housing 1 is divided into a handle 11 and a main body 12; a battery compartment 13 is rotatably installed below the handle 11; the switch button 4 is slidably installed on the handle 11; the display screen 2, the shooting lens 3, and the WiFi module are fixedly installed on the main body 12; the switch button 4 is electrically connected to the display screen 2, the shooting lens 3, and the WiFi module; Among them, the device uses the above - mentioned detection and analysis method to detect the quality of white - feather chicken meat.
[0026] As mentioned above, the above are only specific embodiments of the present invention, but the protection scope of the present invention is not limited thereto. Any person skilled in the art within the technical scope disclosed by the present invention can easily think of changes or substitutions, which should all be covered within the protection scope of the present invention. Therefore, the protection scope of the present invention should be subject to the protection scope of the claims.
Claims
1. A detection and analysis method for a handheld quality detection device for white - feather broilers, characterized in that, The method specifically includes the following steps: S1: Collect photos of the chicken to be detected, store and upload the photos to the detection terminal; S2: Image preprocessing. Gray-scale the image obtained by front-end collection in step S1, and perform weighted averaging on the red, green, and blue components according to the color characteristics of the chicken to obtain a more reasonable gray-scale image, which is convenient for subsequent comparison and matching with the database; S3: Binarize the preprocessed image in step S2 to reduce the difficulty of comparison; S4: Connect the pixels in the regions of the binarized image in step S3 according to the chicken quality evaluation indicators, and obtain the total pixel number of the connected regions under each indicator; S5: Synchronously perform color comparison and analysis on the regions with large color value differences in the evaluation indicators; S6: Summarize the data of the above steps, count the pixel data of each indicator region, and calculate the actual area of each indicator region; S7: Compare the area data obtained in step S6 with the preset data in the database to judge the chicken quality, and send the analysis result to the front end.
2. The detection and analysis method of a handheld white - feather broiler quality detection device according to claim 1, characterized in that, In step S4, the specific steps to obtain the pixel number of the evaluation indicator region image are as follows: According to the gray-scale reference threshold of the white feather chicken root, the congestion gray-scale reference threshold, and the yellow skin gray-scale reference threshold preset in the background, further image processing is performed on the image data binarized in step S3. The specific steps include: S41: Scan the image and select the current pixel point B(x, y)=1; S42: Take B(x, y) as the initial position of the pixel point, assign a label to the pixel point, and then push all adjacent foreground pixels of the pixel point onto the stack; S43: Pop the top pixel of the stack, assign it the same label, and then push all adjacent foreground pixels of the top pixel onto the stack; S44: Repeat step S43 until the stack is empty; S45: Output the results, respectively obtain the gray-scale connected region image of the chicken root, the congestion gray-scale connected region image, and the yellow skin gray-scale connected region image, and calculate the number of pixels in the connected region.
3. The detection and analysis method of a handheld white - feather broiler quality detection device according to claim 2, characterized in that, In step S5, the specific steps of color comparison are as follows: S51: Separate the congestion gray-scale connected region image and the yellow skin gray-scale connected region image obtained in step S45, and respectively obtain the central pixel coordinate values (X1, Y1), (X2, Y2)... (Xn, Yn) of each connected region; S52: Extract 8 pixel points adjacent to the coordinate values with the above coordinate values as the center, and calculate the color difference of the pixels of the 8 pixel points and the central pixel point, a total of 9 points; S53: Quantify the color information of the pixel region obtained in S52, and generate a color feature vector based on the percentage of the total pixels of each color of red, green, and blue. If the percentage is greater than 90%, it reaches the judgment standard.
4. A handheld white - feather broiler quality detection device, comprising a detector housing (1), a display screen (2), a shooting lens (3), a switch button (4), and a WiFi module; the detector housing (1) is divided into a handle (11) and a main body (12); a battery compartment (13) is rotatably installed below the handle (11); the switch button (4) is slidably installed on the handle (11); the display screen (2), the shooting lens (3), and the WiFi module are fixedly installed on the main body (12); the switch button (4) is electrically connected to the display screen (2), the shooting lens (3), and the WiFi module; It is characterized in that This device uses the detection and analysis method of the handheld white - feather broiler quality detection device according to any one of claims 1 - 3 to detect the quality of white - feather chicken meat.