Slaughter line carcass length intelligent real-time measurement method and system based on deep learning

Through the deep learning model YOLOv8Pose and depth camera technology, efficient and accurate measurement of the carcass length of the slaughter line is achieved, solving the problems of low efficiency, high cost and large errors of traditional measurement methods, and providing scientific breeding data and high frequency production support.

CN118967771AInactive Publication Date: 2024-11-15HUAZHONG AGRI UNIV +1

Patent Information

Application Number
CN202411044324.5
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2024-07-31
Publication Date
2024-11-15
Estimated Expiration
Not applicable · inactive patent

AI Technical Summary

Technical Problem

The prior art lacks efficient and accurate automated solutions in the measurement of carcass length in slaughter line, resulting in low measurement efficiency, high cost and prone to errors, making it difficult to meet the needs of large-scale production.

Method used

Using the intelligent real-time measurement method based on the deep learning model YOLOv8Pose, the carcass image is collected through a depth camera and combined with data enhancement technology to achieve accurate measurement of carcass length, and a real-time measurement system including image acquisition, data processing and result display is constructed, using Europause. The distance algorithm calculates the carcass length.

Benefits of technology

It realizes high-precision and real-time carcass length measurement in complex slaughter line environments, improves production efficiency, reduces labor costs, provides scientific breeding data support, adapts to high-frequency production environments, and reduces production costs.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN118967771A_ABST
    Figure CN118967771A_ABST
Patent Text Reader

Abstract

The invention belongs to the field of intelligent agriculture and animal husbandry automation, and particularly relates to a slaughter line carcass length real-time measurement method and system based on a deep learning technology, and the method comprises the steps: collecting three-dimensional data with a depth map in real time through a depth camera, analyzing an image through a deep learning model, and measuring the carcass length; and the distance between the key points is calculated through an Euclidean distance algorithm, so that accurate measurement is realized. The image acquisition module is responsible for acquiring three-dimensional data in real time, the data processing module analyzes images and measures the carcass length by using a deep learning technology, and the result display module displays measurement results in real time, stores and analyzes the images, draws key points and connecting lines, and records the measurement results, so that subsequent data analysis is facilitated. According to the system, the generalization ability of the model in various environments is improved through a data enhancement technology, the frame number of pigs passing through the ROI is recorded through a time sequence counting method, and the accuracy and automatic screening of data are ensured. The system has the advantages of high precision, real-time performance and high efficiency.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present invention belongs to the field of intelligent agriculture and animal husbandry automation, and in particular relates to a real-time measurement method and system for carcass length on a slaughter line based on a deep learning model. Background Art

[0002] With the rapid development of global agriculture and food industry, the slaughtering industry is gradually transforming towards intensification, automation and intelligence. In this context, achieving efficient measurement and management on the slaughtering line has become a key factor in improving production efficiency and product quality. At present, there is no artificial intelligence solution specifically for carcass length measurement on the slaughtering line on the market. Traditional measurement methods, such as manual measurement, are widely used, but still have many shortcomings.

[0003] The manual measurement method is cumbersome, time-consuming and laborious, and the measurement results depend on the workers' experience and skills, which is prone to errors and difficult to meet the needs of large-scale production. In addition, manual measurement is inefficient and cannot meet the requirements of high-frequency production on the slaughter line, which makes the entire measurement process not only time-consuming but also costly.

[0004] On the slaughter line, fast and accurate measurement of carcass length is of great significance for integrating slaughter data to carry out breeding and optimize carcass grading in the entire industry chain. The present invention proposes an intelligent real-time measurement method for carcass length on the slaughter line based on deep learning. By using a deep camera to collect carcass images and combining the efficient real-time reasoning ability of the deep learning model, the system can achieve accurate and fast measurement of carcass length in a complex slaughter line environment. This method has the following advantages and innovations in the field of slaughter line measurement:

[0005] 1. The first application of artificial intelligence in slaughter line measurement: This invention introduces artificial intelligence technology in carcass length measurement, creating a precedent for the application of artificial intelligence in slaughter line measurement. This innovative application not only improves the accuracy and efficiency of measurement, but also realizes full automation and intelligence.

[0006] 2. Full automation and intelligence: Using deep learning technology, the whole process is automated, avoiding the errors and uncertainties of manual operation, and providing more objective and reliable measurement data. The system can automatically capture and analyze carcass images and output measurement results in real time without manual intervention.

[0007] 3. Real-time and high efficiency: The deep learning model used in the present invention has efficient real-time reasoning capabilities and can quickly output measurement results under the high-frequency production environment conditions of the slaughter line, meeting the real-time requirements of the production line. This real-time measurement capability ensures the efficient operation of the production line and avoids production stagnation caused by measurement delays.

[0008] 4. Low cost and high efficiency: Compared with traditional manual and mechanical equipment measurement, the deep learning-based solution is low-cost and easy to deploy. The system does not require expensive mechanical equipment and a large amount of manual operation, which can significantly reduce production costs and improve economic benefits.

[0009] 5. Feedback breeding support: Through the analysis of measurement data, provide scientific basis for breeding, optimize breeding strategies, and improve breeding efficiency and effect. The system can automatically record and analyze the data of each measurement, provide detailed feedback information to breeders, and help scientific breeding.

[0010] In summary, the present invention realizes the intelligent real-time measurement of carcass length on the slaughter line by introducing advanced deep learning technology and efficient system integration, overcomes many shortcomings of traditional methods, has significant innovation and broad application prospects, and will greatly promote the automation and intelligentization of the slaughtering industry. Summary of the invention

[0011] In view of the problems existing in the prior art, the present invention provides a real-time intelligent measurement method and system for carcass length of a slaughter line based on the deep learning model YOLOv8 (YouOnly Look Once). By using the YOLOv8Pose model to predict the key points of the carcass, accurate measurement of the carcass length is achieved. This method aims to overcome the many difficulties of the prior art in automated measurement, provide an efficient and accurate measurement method, and comprehensively improve the production efficiency of the slaughter line. Through deep learning technology and real-time image processing, the system ensures stable and efficient accurate measurement even in complex slaughtering environments, greatly improving production efficiency and reducing labor costs.

[0012] The present invention is implemented as follows: a method for intelligent real-time measurement of carcass length in a slaughter line based on deep learning, comprising the following steps:

[0013] S1, Creation and application of proprietary dataset: Develop a proprietary dataset of 2431 slaughterhouse carcass images. Image acquisition was performed on advanced slaughter lines, covering different enhancement conditions to ensure the diversity and comprehensiveness of the data. The diversity of the dataset ensures that the model can work stably in various complex environments;

[0014] S2, key point annotation of body length: During the image annotation process, after the carcass is segmented, only the inner area of ​​the left fan is annotated. This annotation method can reduce annotation noise and improve the accuracy of prediction. Key point annotation enables the model to accurately identify and locate the key positions of the carcass, thereby more accurately calculating the carcass length;

[0015] S3, data enhancement and generalization capability improvement: Increase the diversity of the data set and improve the generalization capability of the model through data enhancement technology targeting key points. Data enhancement technology includes ten methods such as color transformation and exposure adjustment. Through these technologies, more training samples are generated to improve the performance of the model in practical applications;

[0016] S4, deep learning model selection and training: select and apply the YOLOv8Pose model, which has efficient real-time reasoning capabilities and is particularly suitable for key point detection and pose estimation. The wide range of applications of YOLOv8 makes it not limited to target detection. Through the key point detection function, YOLOv8Pose can play an important role in carcass measurement on the slaughter line. During the training process, determine the appropriate hyperparameters, such as epochs, batch size, etc., to ensure the efficiency and accuracy of the model;

[0017] S5, construction of real-time measurement system: construct a real-time measurement system including image acquisition module, data processing module and result display module. The image acquisition module uses a depth camera to collect three-dimensional data with RGB image and depth image in real time. The data processing module uses the trained YOLOv8Pose model to analyze the image and measure the carcass length. The result display module displays the measurement results in real time, stores the analyzed images, draws key points, key point connections, and marks the distance. The measurement results are recorded in the form of CSV tables and LOG logs to facilitate subsequent review and data analysis;

[0018] S6, body length prediction algorithm: The carcass length is calculated using Euclidean distance. By fusing the depth map and RGB map, the RGB detection points in the video stream are aligned to the corresponding points in the depth map, and the Euclidean distance between the key points of the carcass is calculated using the data in the depth map, so as to accurately measure the carcass length;

[0019] S7, time series counting method: using the characteristics of the slaughter line, by recording the number of frames of pigs detected through the ROI (region of interest), continuous frames represent the same pig, and disconnected frames identify the next pig. This method filters out outliers in data selection to ensure data accuracy and automatic screening;

[0020] Furthermore, in S1, a total of 160 pigs were captured for 15 minutes and 20 seconds. The slaughter line was moving at a uniform speed. The algorithm extracted the video data of the first 9 minutes and 30 seconds (the first 100 pigs) every 2 seconds, extracted 287 images, and selected 221 images. The video stream of the images after 9 minutes and 30 seconds was used as the test set for inference fitting, that is, the data of the last 60 pigs was fitted, and the data of the first 100 pigs was used for training and verification;

[0021] Furthermore, S2 was labeled using the Labelme tool. According to the national standard, two points of the carcass were selected as key points for labeling, namely the front edge of the pubic symphysis and the front edge of the first cervical vertebra. All detection boxes and key point labels were labeled by one person to ensure uniformity and reduce subjective errors. The labeling results were saved in Json format and then converted into YOLO format by the algorithm.

[0022] Furthermore, in S3, data augmentation techniques include: using OpenCV to apply image data augmentation to each training set image that has completed key point annotations. The core of this is to change the visual characteristics of the image without affecting the key point labels to simulate changes in the real world, such as color transformation, exposure adjustment, saturation adjustment, brightness change, and simulated image noise and blur. This method aims to train a robust model that can adapt to different image quality conditions and help the model learn to maintain the stability of key point detection under visual changes.

[0023] Furthermore, S4 divides the labeled data into training set and validation set in a ratio of 10:1. The first 100 data are designated as training set and validation set, and the last 60 data are used as the fitting test set. After the training set and validation set images are labeled and data augmented, model training begins, specifically:

[0024] S41, considering the line speed of the slaughterhouse and the future real-time requirements, the end-to-end YOLOv8 model is selected. The YOLOv8pose posture estimation model can detect key points to meet the needs of carcass length prediction;

[0025] S42, after model selection, the model is trained. The YOLOv8 model is the core of this study, and the ultralytics library is used to train the YOLOv8 model. YOLOv8npose is used as the main model, and the training cycle is set to 500 epochs. As a control, we made a fully trained model, that is, set the early stopping patience of 100 epochs. In these 100 epochs, the training is terminated without any prompt.

[0026] The input image size is 640*640 resolution, and the RTX 3090 GPU is used for training. Other parameters including learning rate and batch size use the default weights of ultralytics for YOLOv8 training. During the training process, data enhancement is also used, including hue, saturation, brightness adjustment intensity, as well as translation, zoom, mosaic data enhancement, etc.

[0027] Furthermore, S5 constructs a real-time measurement system including an image acquisition module, a data processing module and a result display module. The image acquisition module uses a depth camera to collect three-dimensional data with RGB images and depth images in real time. The specific steps are as follows:

[0028] S51, image acquisition module

[0029] The depth camera is used to collect 3D data with RGB and depth images in real time, ensuring the comprehensiveness and accuracy of data collection, and being able to capture the key points and details of the carcass. The depth camera obtains depth map data, enabling the system to accurately measure distance information in 3D space, providing a reliable data basis for subsequent carcass length measurement. The depth camera is ensured to obtain high-quality RGB and depth map data in real time by simultaneously collecting 2D video stream data and depth map stream data, avoiding data alignment issues.

[0030] S52, data processing module

[0031] The trained YOLOv8Pose model is used to analyze and process the collected image data in real time. The model can efficiently detect and predict the key point positions of the carcass, and align and fuse the RGB image with the depth image to ensure the accurate position of each key point in three-dimensional space. The distance between key points is calculated using the Euclidean distance algorithm to accurately measure the carcass length. After detecting the key points, the data processing module immediately calculates and records the carcass length to ensure real-time and efficient measurement.

[0032] S53, result display module

[0033] The measurement results are displayed in real time, including drawing key points, connecting lines between key points, and marking the distances between key points. The display content is clear and intuitive, which is convenient for operators to monitor and check the measurement results in real time. At the same time, the analyzed images are stored for later review. The key points and their connecting lines are marked in the image, and the distances are marked, and label information of the measured carcass length is provided. The measurement results are recorded in the form of CSV tables and logs, including detailed data and timestamps for each measurement, which is convenient for subsequent review and data analysis. The system automatically saves and manages the measurement data, provides historical data query functions, and supports breeding feedback and scientific research;

[0034] S54, Data Recording and Storage

[0035] The measurement results are not only displayed in real time, but also stored as visual images. The key points, key point connections, and distances are clearly drawn in the image. In addition, the system can automatically save and manage measurement data, and the measurement results are recorded in CSV tables and logs, including detailed data and timestamps for each measurement, which is convenient for subsequent review and data analysis.

[0036] Through the above steps, the present invention realizes the intelligent real-time measurement of carcass length in a complex slaughter line environment, and has the advantages of high precision, real-time and high efficiency. The system not only significantly improves the production efficiency of the slaughter line and reduces labor costs, but also provides scientific feedback data for breeding, and has broad application prospects and economic benefits.

[0037] Furthermore, the S6 body length prediction algorithm uses Euclidean distance to calculate the carcass length. The specific steps are as follows:

[0038] S61, alignment fusion of depth stream and RGB stream

[0039] The depth camera is used to simultaneously capture RGB images and depth map images to ensure the synchronization and accuracy of the two. Through the video stream alignment technology, the detection points in the RGB image are accurately matched with the corresponding points in the depth image. The key point positions in the RGB image are detected in real time in the video stream, and the depth information of these key points in the three-dimensional space is obtained through the corresponding depth image. The alignment process ensures the precise matching and consistency of key points in different image sources, thereby ensuring the accurate position of each key point in the three-dimensional space.

[0040] S62, Euclidean distance calculation

[0041] The Euclidean distance between the key points of the carcass is calculated using the data in the depth map. The specific calculation method is:

[0042] Select multiple key points of the carcass, calculate the three-dimensional coordinates of each key point, and obtain its position in three-dimensional space.

[0043] Calculate the Euclidean distance between adjacent key points, the formula is:

[0044]

[0045] The total length of the carcass can be obtained by adding up the Euclidean distances between all adjacent key points.

[0046] Through the above steps, the present invention realizes the accurate measurement of the carcass length of the slaughter line, and has the advantages of high precision and high efficiency. By fusing the data of the RGB image and the depth image, the distance between key points is calculated using the Euclidean distance algorithm, which ensures the accuracy and reliability of the carcass length measurement.

[0047] Furthermore, the S7 time series counting method uses the characteristics of the slaughter line to record the number of frames of pigs detected through the ROI (region of interest). Continuous frames represent the same pig, and disconnected frames identify the next pig. This method filters out outliers in data selection to ensure data accuracy and automatic screening. The specific steps are as follows:

[0048] S71, Region of Interest (ROI) Selection

[0049] The selection of the region of interest is based on the actual application scenario of the slaughter line, and is selected in a stable and stable place on the slaughter line. This can effectively record every pig passing through the area and reduce detection errors caused by movement and vibration. The algorithm determines the best ROI area so that every pig can be accurately recorded.

[0050] S72, frame number recording

[0051] In the selected ROI area, the pigs in each frame are detected through the video stream. The consecutively detected frames represent the same pig. When the frames are disconnected, it marks the beginning of the next pig. The system will record the number of frames for each pig passing through the ROI area to ensure the complete record of each pig.

[0052] S73, outlier filtering

[0053] During the recording process, the system automatically filters out outliers to ensure data accuracy. By setting a reasonable frame number threshold, abnormal frames are filtered out to ensure that the data recording of each pig is accurate. The outlier filtering algorithm will be continuously optimized according to the actual application environment to ensure high data quality.

[0054] S74, Data Recording and Storage

[0055] The system stores the data of each pig detected and recorded, and generates a detailed measurement report. The data storage includes the frame number record of the pig, the timestamp of the pig passing through the ROI area, the measured carcass length and other information. These data will be recorded in the form of CSV tables and logs to facilitate subsequent review and data analysis.

[0056] Through the above steps, the time series counting method of the present invention can effectively utilize the characteristics of the slaughter line, accurately record the number of frames of each pig, filter outliers, ensure the accuracy and automatic screening of data. This method not only improves the accuracy and reliability of data recording, but also supports detailed data analysis and breeding research, and has broad application prospects and economic benefits.

[0057] Another object of the present invention is to provide a system for applying the deep learning-based intelligent real-time measurement method for slaughter line carcass length, comprising:

[0058] Image acquisition module: used to use the depth camera to collect three-dimensional data with depth map in real time to ensure the comprehensiveness and accuracy of data acquisition;

[0059] Data processing module: used to use the trained YOLOv8Pose model to analyze image data, predict key points of body length, and measure carcass length. After detecting the key points, the data processing module immediately calculates and records the carcass length, and calculates the distance between key points through the Euclidean distance algorithm to achieve accurate measurement;

[0060] Result display module: used to display the measurement results in real time, including drawing key points, connecting lines between key points, and marking the distance between key points. This module also stores the analyzed images, and the measurement results are recorded in CSV tables and logs to facilitate subsequent review and data analysis;

[0061] Data recording and storage module: used to store the measurement results as visual images, in which key points, key point connections, and distances are clearly drawn. In addition, the measurement results are recorded in CSV tables and logs, including detailed data and timestamps for each measurement, to facilitate subsequent review and data analysis;

[0062] Another object of the present invention is to provide a computer device, which includes a memory and a processor, wherein the memory stores a computer program, and when the computer program is executed by the processor, the processor executes the steps of the intelligent real-time measurement method for slaughter line carcass length based on deep learning.

[0063] Another object of the present invention is to provide a computer-readable storage medium storing a computer program. When the computer program is executed by a processor, the processor executes the steps of the intelligent real-time measurement method for slaughter line carcass length based on deep learning.

[0064] Another object of the present invention is to provide an information data processing terminal, which is used to implement the deep learning-based intelligent real-time measurement system for slaughter line carcass length.

[0065] In combination with the above technical solutions and the technical problems solved, the advantages and positive effects of the technical solutions to be protected by the present invention are as follows:

[0066] First, in view of the technical problems existing in the above-mentioned prior art, some creative technical effects are brought about after solving the problems. The specific description is as follows:

[0067] (1) Improve measurement accuracy and efficiency: By introducing artificial intelligence technology and deep learning models, high-precision automated measurement is achieved, solving the problems of large errors and low efficiency in manual measurement. The deep learning model has efficient real-time reasoning capabilities and can quickly output measurement results under the high-frequency production environment conditions of the slaughter line to ensure efficient operation of the production line. This real-time measurement capability avoids production stagnation caused by measurement delays and improves production efficiency.

[0068] (2) Full automation and intelligence: Deep learning technology is used to achieve full automation of measurement, avoiding errors and uncertainties in manual operation and providing more objective and reliable measurement data. The system can automatically capture and analyze carcass images and output measurement results in real time without manual intervention, greatly reducing labor costs. At the same time, automated measurement reduces dependence on manual experience and skills and improves data consistency and reliability.

[0069] (3) Enhance the generalization ability of the model: By creating a proprietary dataset containing 2,431 carcass images, covering different enhancement conditions, we ensure the diversity and comprehensiveness of the data. The data enhancement technology includes ten methods such as color transformation, exposure adjustment, and saturation adjustment, which generate more training samples and improve the performance of the model in complex environments. The diversity of the dataset ensures that the model can work stably in various complex environments and provide highly accurate measurement results.

[0070] (4) Reduce production costs and improve economic benefits: Compared with traditional manual and mechanical equipment measurement, the deep learning-based solution is low-cost and easy to deploy. The system does not require expensive mechanical equipment and a large amount of manual operation, which can significantly reduce production costs and improve economic benefits. The system's high efficiency and low cost make it highly cost-effective and suitable for promotion and application in large-scale production environments.

[0071] (5) Support scientific breeding and optimize carcass grading: Through the analysis of measurement data, provide a scientific basis for breeding, optimize breeding strategies, and improve breeding efficiency and effectiveness. The system can automatically record and analyze the data of each measurement, provide detailed feedback information to farmers, and assist scientific breeding. These data can also be used to optimize carcass grading, provide more accurate grading standards, and improve product quality.

[0072] (6) Ensure the accuracy and stability of data: By using the body length key point annotation technology, only the inner area of ​​the left fan is annotated during the image annotation process, which reduces the annotation noise and improves the accuracy of the prediction. Key point annotation enables the model to accurately identify and locate the key positions of the carcass, thereby more accurately calculating the carcass length. By fusing the depth map and the RGB map, the RGB detection points in the video stream are aligned to the corresponding points in the depth map, and the Euclidean distance between the key points of the carcass is calculated using the data in the depth map, ensuring the accuracy and stability of the measurement.

[0073] (7) Convenient for grading and management: The system can output measurement results in real time, including drawing key points, connecting lines between key points, and marking the distances between key points. The measurement results are recorded in CSV tables and LOG logs to facilitate subsequent review and data analysis. The system automatically saves and manages measurement data, provides historical data query functions, and supports breeding feedback and scientific research. These functions facilitate accurate grading and management of the slaughter line and improve management efficiency.

[0074] In summary, the present invention overcomes many shortcomings of traditional methods by introducing advanced deep learning technology and efficient system integration, and realizes intelligent real-time measurement of carcass length on the slaughter line. It has significant innovation and broad application prospects, and will greatly promote the automation and intelligentization of the slaughtering industry.

[0075] Second, considering the technical solution as a whole or from the perspective of the product, the technical effects and advantages of the technical solution to be protected by the present invention are described in detail as follows:

[0076] This invention represents a major advancement in automated management of the slaughtering industry. By adopting the latest computer vision and deep learning technologies, this invention not only improves the accuracy and efficiency of carcass length measurement, but also significantly improves the stability and generalization ability of the model under different environmental conditions. The system is able to work stably in a complex slaughtering line environment, reduce errors and deviations caused by manual operation, and achieve high consistency and reliability of measurement results. This innovative technology greatly improves production efficiency, avoids production delays and stagnation, and ensures efficient operation of the slaughtering line. Compared with traditional manual and mechanical measurement methods, this system has the characteristics of low deployment cost and simple maintenance, significantly reducing the overall operating cost and reducing dependence on labor.

[0077] (1) Accurate measurement: By using the deep learning model YOLOv8 and key point detection technology, the present invention can achieve high-precision measurement of the length of the carcass on the slaughter line. By fusion of the depth map and RGB map, the detection points in the video stream are aligned, and the Euclidean distance between the key points of the carcass is calculated to ensure the accuracy of the measurement results. Compared with traditional manual measurement methods, the system provides more consistent and reliable data.

[0078] (2) High efficiency and real-time: The system of the present invention has high efficiency and real-time reasoning capability, and can quickly output measurement results in the high-frequency production environment of the slaughter line. The system collects three-dimensional data in real time through a depth camera, and uses the trained YOLOv8Pose model for real-time analysis and processing, ensuring the real-time and high efficiency of the measurement and avoiding delays in the production process.

[0079] (3) Fully automated and intelligent: The system achieves fully automated measurement without manual intervention, reducing manual operation errors and uncertainties. The automated and intelligent measurement process not only improves the objectivity and reliability of data, but also significantly reduces labor costs, adapting to the needs of the modern slaughtering industry's transformation to intensive, automated, and intelligent production.

[0080] (4) Enhanced generalization capability: This paper improves the generalization capability of the model through data augmentation technology and diversified proprietary data sets. The data sets cover different environments and conditions, and more training samples are generated through color transformation, exposure adjustment and other technologies to ensure that the model works stably in various complex environments and provides highly accurate measurement results.

[0081] (5) Low cost and high efficiency: Compared with traditional mechanical equipment and manual measurement methods, solutions based on deep learning are low cost and easy to deploy. The system does not require expensive mechanical equipment and a large amount of manual operation, which can significantly reduce production costs and improve economic benefits. This cost-effective technical solution is suitable for promotion and application in large-scale production environments.

[0082] (6) Support scientific breeding and optimized grading: The system automatically records and analyzes the data of each measurement, provides a scientific basis for breeding, optimizes breeding strategies, and improves breeding efficiency and results. At the same time, the measurement data supports the accurate grading of carcasses, provides more accurate grading standards, improves product quality, and provides data support for the management and decision-making of the slaughtering industry.

[0083] (7) Data management and visualization: The system outputs measurement results in real time, including key point drawing, key point connection and distance annotation. The results are recorded in CSV form and LOG log to facilitate subsequent review and data analysis. The system automatically saves and manages measurement data, provides historical data query function, supports breeding feedback and scientific research, and improves the convenience and effectiveness of data management.

[0084] (8) Reliability and stability: Through the creation of proprietary data sets, body length key point annotation and data enhancement technology, the system is ensured to work stably in various complex environments. Efficient data processing and accurate measurement algorithms ensure the reliability and stability of the measurement results, meeting the strict requirements of high-frequency production of the slaughter line.

[0085] In summary, the present invention provides a high-precision, high-efficiency, low-cost intelligent real-time measurement method and system for carcass length on the slaughter line through deep learning technology, real-time image processing and intelligent system integration, which overcomes many shortcomings of traditional methods, has significant technical effects and broad application prospects, and promotes the development of the slaughtering industry towards automation and intelligence.

[0086] Third, as auxiliary evidence of the inventiveness of the claims of the present invention, it is also reflected in the following important aspects:

[0087] (1) The expected benefits and commercial value of the technical solution of the present invention after transformation are:

[0088] The intelligent carcass length measurement system of the present invention has significant commercial value and expected benefits, especially in improving the production efficiency and management level of slaughterhouses. Through accurate carcass length measurement technology, the management efficiency of the slaughter line is greatly improved, allowing operators to perform more accurate production monitoring, product optimization and quality control, thereby improving overall production efficiency and output quality. In addition, this automated measurement method reduces reliance on manual inspections, reduces labor costs, and significantly reduces problems caused by human errors, such as subjective errors of measurement personnel and inconsistent measurement standards.

[0089] (1) Improved production efficiency: By realizing intelligent real-time measurement of carcass length at the slaughter line, the present invention greatly improves the efficiency of the measurement process. Compared with traditional manual measurement methods, the system can quickly output accurate measurement results in a high-frequency production environment, reducing delays in the production process and significantly improving overall production efficiency.

[0090] (2) Reduce labor costs: The present invention realizes full automation and intelligent measurement, reduces reliance on manual operation, and reduces labor costs. Enterprises do not need to invest a large amount of manpower to measure carcass length, thereby saving human resource costs and reducing additional costs caused by manual errors.

[0091] (3) Improve measurement accuracy: By using the deep learning model YOLOv8 and key point detection technology, the present invention can provide high-precision measurement results. Compared with manual measurement methods, the system reduces the errors caused by human factors, improves the accuracy and consistency of data, and provides reliable data support for quality control in the production process.

[0092] (4) Support for refined management: The system records and analyzes measurement data in real time, providing detailed production information for slaughtering enterprises. Enterprises can conduct refined management and decision-making optimization based on these data, improve management efficiency and production quality. Data visualization and historical data query functions enable managers to better monitor and optimize production processes.

[0093] (5) Enhance market competitiveness: By introducing advanced deep learning technology and intelligent measurement systems, enterprises can significantly improve their technical level and production capacity and enhance their market competitiveness. As the slaughtering industry is transforming towards intensification, automation and intelligence, the application of this invention will enable enterprises to occupy an advantageous position in the market.

[0094] (6) Optimizing breeding strategies: The precise measurement data provided by the system provides a reliable basis for scientific breeding, helping breeding companies optimize breeding strategies and improve breeding efficiency and effectiveness. By analyzing the measurement data, companies can better select excellent varieties and improve breeding methods, thereby increasing production efficiency.

[0095] (7) Promote industrial upgrading: The application of the present invention will promote the development of the slaughtering industry towards automation and intelligence, and promote industrial upgrading. By introducing advanced technologies, enterprises can improve production technology levels, optimize production processes, and promote technological progress and innovative development of the entire industry.

[0096] (8) Reduce equipment maintenance costs: The system uses a deep learning-based solution to reduce reliance on traditional mechanical equipment. Compared with expensive and complex mechanical equipment, the intelligent measurement system is easy to maintain, which reduces equipment maintenance costs and reduces the impact of equipment failures on production.

[0097] (9) Improve customer satisfaction: By achieving efficient and accurate measurement and optimized production management, companies can provide higher quality products and improve customer satisfaction. Accurate carcass length data helps optimize product grading and pricing strategies and meet customers' high demands for product quality.

[0098] In summary, the present invention provides a high-precision, high-efficiency, low-cost intelligent real-time measurement method and system for carcass length on the slaughter line through deep learning technology, real-time image processing and intelligent system integration, which overcomes many shortcomings of traditional methods, has significant technical effects and broad application prospects, and promotes the development of the slaughtering industry towards automation and intelligence.

[0099] (2) The technical solution of the present invention fills the technical gap in the industry at home and abroad:

[0100] The intelligent pig carcass length measurement system proposed in this invention represents an important breakthrough in the field of animal husbandry technology at home and abroad, filling the technical gap in carcass length measurement on the slaughter line. The system uses advanced deep learning technology, combined with unique data enhancement and time series counting methods, to provide an innovative solution for the precise measurement and management of carcasses.

[0101] The high accuracy and reliability of this system are due to the deep learning model YOLOv8Pose and the image data set unique to the slaughter line, which can achieve high measurement accuracy. Through continuous learning and optimization, the system can adapt to various slaughtering environments and different breeds of pigs to ensure accurate measurement of carcass length under various conditions. In addition, the system has powerful data processing capabilities. It can not only process complex image data, but also extract key features from it for precise measurement. At the same time, the data provided supports decision-making, optimizes production processes and management strategies. It integrates the depth image and two-dimensional color image in the depth camera to achieve synchronous processing of image and depth data, ensuring the accuracy and real-time nature of the data.

[0102] In terms of feedback breeding support, the system can record and analyze the data of each measurement, provide detailed feedback information to breeders, help optimize breeding strategies, and improve breeding efficiency and effectiveness. Through data-driven decision support, companies can better understand the key factors in the breeding process and promote the development of scientific breeding.

[0103] This system has a wide range of application potential. Although it is mainly used for measuring pig carcass length, its technical framework and methods have broad application prospects. The modular design and flexibility of the system enable it to adapt to different application scenarios and needs, providing a variety of possibilities for automated management of animal husbandry. Through the above advantages and innovations, the present invention fills the gaps in the industry at home and abroad in terms of technology, provides an efficient, accurate and reliable method for measuring carcass length, and promotes technological progress and management optimization in the slaughtering industry. The wide application of the system can not only improve production efficiency and product quality, but also bring significant economic benefits and market competitiveness to breeding enterprises, and at the same time provide valuable technical support and inspiration for research and application in related fields.

[0104] Specifically, the present invention has significant innovations and advantages in the following aspects:

[0105] The first application of artificial intelligence in slaughter line measurement: This invention introduces the deep learning model YOLOv8 into the slaughter line carcass length measurement, realizing full automation and intelligent measurement. This is the first time that artificial intelligence technology has been applied to slaughter line measurement at home and abroad, creating a precedent for the application of artificial intelligence in this field and filling the technical gap in this field.

[0106] Breaking through the limitations of traditional measurement methods: Traditional slaughter line carcass length measurement methods mainly rely on manual operation, which has problems such as cumbersome operation, time-consuming and labor-intensive, and large errors. This invention overcomes these shortcomings through deep learning and image processing technology, realizes efficient and accurate real-time measurement, and significantly improves the accuracy and efficiency of measurement.

[0107] Innovative data enhancement and generalization technology: This paper uses a variety of data enhancement technologies such as color transformation and exposure adjustment to create a proprietary dataset containing 2,431 carcass images. The diversity of these technologies and datasets ensures that the model can work stably in various complex environments, significantly improves the generalization ability of the model, and fills the technical gap of stable measurement in complex slaughtering environments.

[0108] Combination of real-time and high efficiency: The present invention utilizes the efficient real-time reasoning capability of the deep learning model to quickly output measurement results under the high-frequency production environment of the slaughter line. This combination of real-time and high efficiency has not been achieved in the prior art, filling the technical gap of real-time measurement under high-frequency production conditions of the slaughter line.

[0109] Automated data recording and analysis: The system can automatically record and analyze measurement data, provide detailed feedback information, support scientific breeding and accurate grading. Compared with traditional methods, this invention greatly simplifies the data management and analysis process, filling the technical gap of automated data recording and analysis.

[0110] Multifunctional integrated system architecture: The present invention constructs a real-time measurement system including image acquisition, data processing and result display modules, realizing multifunctional integration. Traditional measurement systems cannot provide such a comprehensive and integrated solution, and the system architecture of the present invention fills the technical gap in this field.

[0111] Cost-effective solution: Deep learning-based measurement systems are low-cost and easy to deploy, avoiding the need for expensive mechanical equipment and a lot of manual operations. This cost-effective solution has not yet been realized in existing technologies, filling the gap in cost-effective intelligent measurement technology.

[0112] Supporting optimization of the entire industry chain: The system of the present invention can provide accurate measurement data to support breeding optimization and carcass grading of the entire industry chain. Compared with traditional methods, the present invention can better integrate and utilize measurement data, filling the technical gap of optimizing the entire industry chain.

[0113] In summary, the present invention innovatively solves many shortcomings of traditional slaughter line measurement methods by introducing deep learning and image processing technology, fills the technical gap in the intelligent real-time measurement of carcass length on the slaughter line at home and abroad, and has important technical significance and application value.

[0114] (3) The technical solution of the present invention solves the technical problems that people have been eager to solve but have never been able to solve successfully:

[0115] The intelligent carcass length measurement system of the present invention solves many long-standing technical problems in the automated measurement of slaughter lines, and meets the urgent demand for efficient, accurate, and non-invasive measurement technology in the industry. Specifically, the present invention overcomes the following major problems:

[0116] Accuracy of automated measurement: Traditional mechanical measurement methods rely on manual operation, which is not only inefficient but also prone to errors and inconsistencies. This invention uses the deep learning model YOLOv8Pose, through high-precision image recognition and key point detection technology, to achieve accurate measurement of carcass length, completely solving the problems of errors and inefficiency in manual measurement.

[0117] Measurement stability in complex environments: The slaughter line environment is complex, and factors such as lighting conditions and operating speed will affect the measurement results. Traditional methods are difficult to maintain stability in various complex environments. This invention improves the generalization ability of the model in different environments and conditions through the application of data enhancement and deep learning technology, ensuring stable and accurate measurement in various complex environments.

[0118] Real-time and high efficiency: Traditional measurement methods require manual operation, which makes it difficult to achieve real-time measurement and data processing. This invention uses a depth camera combined with a deep learning model to achieve real-time data acquisition, processing and display. The measurement results can be displayed in real time, and automatically stored and analyzed, greatly improving the measurement efficiency and the timeliness of data processing. The high efficiency of the system enables the production line to run seamlessly, avoiding production stagnation due to measurement delays.

[0119] Data-driven decision support: Traditional measurement methods are difficult to provide sufficient data to support decision making. The present invention provides rich decision support information by automatically collecting and analyzing a large amount of measurement data. Through data analysis, breeding companies can optimize production processes, improve quality control, predict market demand, and achieve scientific and data-based management.

[0120] Stable measurement in complex environments: Due to the complex environment of the slaughter line, large changes in lighting and high noise, traditional measurement methods are difficult to ensure the accuracy and stability of measurements in these environments. Despite various attempts in the industry, the results have been unsatisfactory. This invention significantly improves the generalization ability of the model by creating a diverse proprietary data set and applying data enhancement technology, ensuring stable operation in various complex environments, and successfully solving this technical problem.

[0121] Low-cost and high-efficiency solution: Traditional measurement methods rely on expensive mechanical equipment and a lot of manual operations, which are costly and inefficient. Despite the market demand, no cost-effective solution has been found. The deep learning-based measurement system of the present invention has the characteristics of low cost and easy deployment, successfully providing a low-cost and high-efficiency solution to meet market demand.

[0122] Feedback breeding support: The system can record and analyze the data of each measurement, provide detailed feedback information to breeders, help optimize breeding strategies, and improve breeding efficiency and effectiveness. Through data-driven decision support, companies can better understand the key factors in the breeding process and promote the development of scientific breeding. The system not only helps breeders select high-quality breeding individuals, but also allows breeders to improve and perfect breeding plans through historical data analysis and improve the overall breeding level.

[0123] The present invention solves many long-standing technical problems in the automated measurement of slaughter lines, and meets the industry's urgent need for efficient, accurate, and non-invasive measurement technology. Through the application of the present invention, slaughtering companies can not only achieve efficient and accurate carcass length measurement, improve production efficiency and product quality, but also improve management level and market competitiveness through data-driven decision support. This technological breakthrough provides strong technical support for the modern management of the slaughtering industry and even the entire animal husbandry industry, and has broad application prospects and significant economic benefits.

[0124] (4) The technical solution of the present invention overcomes technical prejudice:

[0125] The intelligent carcass length measurement system of the present invention overcomes various technical biases in traditional slaughter line measurement methods and breaks through the long-standing mindset and technical limitations in the industry, which is specifically reflected in the following aspects:

[0126] Transition from mechanical measurement to intelligent measurement: Traditional slaughter line measurement methods rely on mechanical equipment and manual operation, which have problems such as low efficiency, poor accuracy, and easy errors. Mechanical measurement methods are considered to be the only reliable solution, thus forming a technical bias. This invention successfully realizes intelligent and non-invasive carcass length measurement by introducing deep learning models and image processing technology, greatly improving measurement accuracy and efficiency, and changing the industry's inherent dependence on mechanical measurement methods.

[0127] Breaking through the reliance on manual measurement: Traditional slaughter line measurement methods are highly dependent on manual operation. Although manual measurement has errors and low efficiency, it has always been considered indispensable due to the limitations of technical means. This invention successfully realizes full-process automated measurement through automated image recognition and data processing technology, reduces the impact of human factors, significantly improves the accuracy and efficiency of measurement, and breaks through the technical prejudice against manual measurement.

[0128] Technical bias of cost and benefit: Traditionally, the slaughter industry believes that high-precision measurement systems must rely on expensive mechanical equipment and a lot of manual operations, resulting in high costs and low benefits. This technical bias limits the development of cost-effective measurement solutions. The present invention provides a low-cost and high-efficiency measurement system based on deep learning, which does not require expensive equipment and a lot of manual operations, significantly reduces production costs, improves economic benefits, and successfully overcomes this technical bias.

[0129] Technical bias against real-time measurement: Many people believe that it is very difficult to achieve real-time measurement in the high-frequency production environment of the slaughter line, and traditional methods are difficult to meet the real-time requirements. This technical bias leads to low expectations for real-time measurement technology. This invention uses the efficient real-time reasoning ability of the deep learning model to successfully achieve real-time measurement under the high-frequency production conditions of the slaughter line, meet the real-time requirements of the production line, and overcome the technical bias against real-time measurement.

[0130] Technical bias towards full automation: There is a common technical bias in the slaughtering industry, which is that full automation and intelligent measurement systems are difficult to achieve and must rely on manual intervention. The present invention achieves full automation and intelligent measurement through system integration of image acquisition, data processing and result display modules, avoids manual operation errors and uncertainties, greatly improves production efficiency and data reliability, and successfully overcomes this technical bias.

[0131] Changing doubts about adaptability to complex environments: Traditional mechanical and manual measurement methods are difficult to maintain stability and efficiency in complex slaughter line environments. The industry generally believes that automated measurement technology is difficult to adapt to complex and changeable actual operating environments. This invention successfully achieves stable and accurate measurement in various complex environments through data enhancement technology and the high generalization ability of deep learning models, changing the industry's doubts and prejudices about the adaptability of automated measurement technology.

[0132] Innovation in data-driven decision-making: Traditional measurement methods often lack systematic data recording and analysis, and cannot provide effective support for management decisions. Data-driven decision-making has not yet been fully valued and applied in the slaughtering industry. This invention provides rich data support for management decisions through automated data collection and analysis, promotes the popularization and application of data-driven decision-making, and breaks the limitations of traditional management models.

[0133] Feedback breeding support: The system can record and analyze the data of each measurement, provide detailed feedback information to breeders, help optimize breeding strategies, and improve breeding efficiency and effectiveness. Through data-driven decision support, companies can better understand the key factors in the breeding process and promote the development of scientific breeding. This data analysis capability has changed the reliance on experience and intuition in the traditional breeding process, and significantly improved the scientificity and efficiency of breeding work.

[0134] In summary, the present invention overcomes a variety of technical prejudices and breaks through the long-standing technical limitations in the industry through an intelligent, non-invasive, efficient and accurate carcass length measurement method. Through the application of the present invention, slaughtering enterprises can not only achieve more efficient and accurate production management, improve product quality and market competitiveness, but also achieve scientific and modern management through data-driven decision support, providing strong technical support and innovative power for the development of the slaughtering industry and even the entire animal husbandry industry. BRIEF DESCRIPTION OF THE DRAWINGS

[0135] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the following briefly introduces the drawings required for use in the embodiments of the present invention. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.

[0136] Figure 1 It is a flow chart of a method for intelligently measuring carcass length on a slaughter line based on deep learning provided by an embodiment of the present invention;

[0137] Figure 2 It is a structural diagram of a slaughter line carcass length intelligent measurement system based on deep learning provided by an embodiment of the present invention;

[0138] Figure 3 This is a diagram of the data set division of the slaughter line carcass length intelligent measurement method based on deep learning. This figure shows the data set division and processing flow in the slaughter line carcass length intelligent measurement method based on deep learning. There are a total of 160 pigs' data, of which 100 pigs' image data are used for training sets and validation sets, and are used for model training after data enhancement. The ratio of training set to validation set is 10:1. The data of 60 pigs is used as a test set for model testing and evaluation.

[0139] Figure 4This is a diagram of the annotation and reasoning of the intelligent measurement method for carcass length at the slaughter line based on deep learning. This diagram shows the schematic process of training set annotation and video stream reasoning in the intelligent measurement method for carcass length at the slaughter line based on deep learning. The left image annotates the key point positions in the training set, and the right image shows the body length measured by the key points during the video stream reasoning process.

[0140] Figure 5 This is a data enhancement diagram of the intelligent measurement method for carcass length at the slaughter line based on deep learning. The figure shows the effect of the original image data after being processed by various data enhancement techniques. After the original image is processed, the data set is enhanced by 11 times. The figure shows different enhancement techniques, including color adjustment, exposure change, noise addition, blur processing, saturation adjustment, and brightness change.

[0141] Figure 6 This is a structural diagram of the slaughter line carcass length intelligent measurement system based on YOLOv8. The diagram shows the structure and workflow of the slaughter line carcass length intelligent measurement system based on the YOLOv8 model. The system processes the input original image through the YOLOv8 model, outputs the key point detection results, and measures and draws the carcass length according to the key points.

[0142] Figure 7 This is a model training effect evaluation diagram of the present invention, which shows the training effect evaluation results of the deep learning model YOLOv8 in the carcass length intelligent measurement task. The figure includes a confusion matrix and an F1Confidence curve, which are used to evaluate the classification accuracy and confidence of the model.

[0143] Figure 8 : is a schematic diagram of carcass length measurement results provided by an embodiment of the present invention, which shows the measurement results of the slaughter line carcass length intelligent measurement system based on the deep learning model YOLOv8. The figure shows the key points detected by the model, the key point connection lines, and the measured carcass length.

[0144] Fig. 9 This is a schematic diagram of the time series counting method, which shows the application process of the time series counting method in the carcass measurement of the slaughter line. By recording the number of pig frames detected through the ROI (region of interest), continuous frames represent the same pig, and the disconnection of the frame number identifies the next pig. This method ensures the accuracy and automatic screening of the data.

[0145] Fig.10 This is a fitting analysis chart of carcass length measurement results and true values. This chart shows the fitting relationship between carcass length measurement results and true values. The left chart shows the correlation analysis between the measured values ​​and true values, and the right chart shows the sorting correlation analysis between the measured values ​​and true values. The relationship between the predicted value and the true value is shown in the figure through the fitting line. DETAILED DESCRIPTION

[0146] In order to make the purpose, technical solution and advantages of the present invention more clearly understood, the present invention is further described in detail below in conjunction with the embodiments. 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.

[0147] In view of the problems existing in the prior art, the present invention provides a method and system for intelligent real-time measurement of carcass length on a slaughter line based on deep learning.

[0148] First, a depth camera and bracket are set up on the slaughter line to ensure that the depth camera can stably and accurately capture the key parts of the carcass. The depth camera collects the RGB image and depth image of the carcass in real time, providing three-dimensional data to ensure the comprehensiveness and accuracy of data collection. The depth camera equipment installed on the slaughter line can continuously and stably obtain the image data of the carcass.

[0149] Next, a proprietary model based on the YOLOv8 model is used. This model has efficient real-time reasoning capabilities and can accurately detect and predict the key point locations of the carcass. The image data collected by the depth camera is input into the trained YOLOv8Pose model, which analyzes and processes the image in real time to detect the key point locations of the carcass. During the image processing and analysis process, the model ensures that each key point is accurately located in three-dimensional space by fusing the data of the RGB image and the depth image.

[0150] The system uses the Euclidean distance algorithm to calculate the distance between key points of the carcass, thereby accurately measuring the length of the carcass. During the measurement process, the system will display the measurement results in real time, such as Figure 8 As shown in the figure, the body length information is displayed and key points and key point lines are drawn. The display content is clear and intuitive, which is convenient for operators to monitor and check the measurement results in real time.

[0151] The system automatically generates logs and tabular data to record the results of each measurement in detail, including carcass length, key point locations and distances. These data are stored in CSV tables and logs for easy review and data analysis. During the measurement process, the system also records the complete drawing result video, saving all image data and measurement results during the measurement process.

[0152] Through the above steps, the present invention realizes the intelligent real-time measurement of carcass length in the slaughter line environment. The system can not only measure the carcass length in real time with high precision, but also automatically generate detailed logs and data records, provide scientific feedback data for production management and breeding, significantly improve the production efficiency of the slaughter line, reduce labor costs, and has broad application prospects and economic benefits.

[0153] like Figure 1 As shown, the method and system for intelligent real-time measurement of carcass length on a slaughter line based on deep learning provided by an embodiment of the present invention include:

[0154] In view of the problems existing in the prior art, the present invention provides a method and system for intelligent real-time measurement of carcass length in a slaughter line based on the deep learning model YOLOv8. By using the YOLOv8Pose model to predict the key points of the carcass, accurate measurement of the carcass length is achieved. This method aims to solve the difficulties of the prior art in automated measurement, provide an efficient and accurate measurement method, and comprehensively improve the production efficiency and product quality of the slaughter line.

[0155] The present invention is implemented as follows: a method for intelligent real-time measurement of carcass length in a slaughter line based on deep learning, comprising the following steps:

[0156] S1, Creation and application of proprietary dataset: Develop a proprietary dataset of 2431 slaughterhouse carcass images. Image acquisition was performed on advanced slaughter lines, covering different enhancement conditions to ensure the diversity and comprehensiveness of the data. The diversity of the dataset ensures that the model can work stably in various complex environments;

[0157] S2, body length key point annotation: data enhancement and generalization ability improvement: in the image annotation process, after the carcass is segmented, only the inner area of ​​the left fan is annotated. This annotation method can reduce annotation noise and improve the accuracy of prediction. Key point annotation enables the model to accurately identify and locate the key positions of the carcass, thereby more accurately calculating the carcass length;

[0158] S3, data enhancement and generalization capability improvement: Increase the diversity of the data set and improve the generalization capability of the model through data enhancement technology targeting key points. Data enhancement technology includes ten methods such as color transformation and exposure adjustment. Through these technologies, more training samples are generated to improve the performance of the model in practical applications;

[0159] S4, deep learning model selection and training: select and apply the YOLOv8Pose model, which has efficient real-time reasoning capabilities and is particularly suitable for key point detection and pose estimation. The wide range of applications of YOLOv8 makes it not limited to target detection. Through the key point detection function, YOLOv8Pose can play an important role in carcass measurement on the slaughter line. During the training process, determine the appropriate hyperparameters, such as epochs, batch size, etc., to ensure the efficiency and accuracy of the model;

[0160] S5, construction of real-time measurement system: construct a real-time measurement system including image acquisition module, data processing module and result display module. The image acquisition module uses a depth camera to collect three-dimensional data with depth map in real time. The data processing module uses the trained YOLOv8Pose model to analyze the image and measure the carcass length. The result display module displays the measurement results in real time, stores the analyzed images, draws key points, key point connections, and marks the distance. The measurement results are recorded in the form of CSV tables and LOG logs to facilitate subsequent review and data analysis;

[0161] S6, body length prediction algorithm: The carcass length is calculated using Euclidean distance. By fusing the depth map and RGB map, the RGB detection points in the video stream are aligned to the corresponding points in the depth map, and the Euclidean distance between the key points of the carcass is calculated using the data in the depth map, so as to accurately measure the carcass length;

[0162] S7, time series counting method: using the characteristics of the slaughter line, by recording the number of frames of pigs detected through the ROI (region of interest), continuous frames represent the same pig, and disconnected frames identify the next pig. This method filters out outliers in data selection to ensure data accuracy and automatic screening.

[0163] Furthermore, in S1, a total of 160 pigs were collected for 15 minutes and 20 seconds. The slaughter line was moving at a constant speed. The algorithm extracted the video data of the first 9 minutes and 30 seconds (the first 100 pigs) every 2 seconds, extracted 287 images, and selected 221 images. The video stream after 9 minutes and 30 seconds was used as the test set for inference fitting, that is, the data of the last 60 pigs was fitted, and the data of the first 100 pigs was used for training and verification.

[0164] Furthermore, S2 was labeled using the Labelme tool. According to the national standard, two points of the carcass were selected as key points for labeling, namely the front edge of the pubic symphysis and the front edge of the first cervical vertebra. All detection boxes and key point labels were labeled by one person to ensure uniformity and reduce subjective errors. The labeling results were saved in Json format and then converted into YOLO format by the algorithm.

[0165] Furthermore, in S3, data augmentation techniques include: using OpenCV to apply image data augmentation to each training set image that has been labeled with key points. The core of this is to change the visual characteristics of the image without affecting the labels of the key points to simulate changes in the real world, such as color transformation, exposure adjustment, saturation adjustment, brightness change, and simulated image noise and blur. This method aims to train a robust model that can adapt to different image quality conditions and help the model learn to maintain the stability of key point detection under visual changes.

[0166] Furthermore, S4 divides the labeled data into training set and validation set in a ratio of 10:1. The first 100 data are designated as training set and validation set, and the last 60 data are used as the fitting test set. After the training set and validation set images are labeled and data augmented, model training begins, specifically:

[0167] S41, considering the line speed of the slaughterhouse and the future real-time requirements, the end-to-end YOLOv8 model is selected. The YOLOv8pose posture estimation model can detect key points to meet the needs of carcass length prediction;

[0168] S42, after model selection, the model is trained. The YOLOv8 model is the core of this study. The ultralytics library is used to train the YOLOv8 model. YOLOv8n is used as the main model. The training cycle is set to 500 epochs, which is medium, not too many, not too few. As a control, we made a fully trained model, that is, set the early stopping patience of 100 epochs. In these 100 epochs, no prompts will be given before the training is terminated.

[0169] The input image size is 640*640 resolution, and the RTX 3090 GPU is used for training. Other parameters including learning rate and batch size use the default weights of ultralytics for YOLOv8 training. During the training process, data enhancement is also used, including hue, saturation, brightness adjustment intensity, as well as translation, zoom, mosaic data enhancement, etc.

[0170] Furthermore, S5 constructs a real-time measurement system including an image acquisition module, a data processing module and a result display module. The image acquisition module uses a depth camera to collect three-dimensional data with a depth map in real time. The specific steps are as follows:

[0171] S51, image acquisition module

[0172] The depth camera is used to collect three-dimensional data with depth map in real time, ensuring the comprehensiveness and accuracy of data collection, and can capture the key points and details of the carcass. The depth map data obtained by the depth camera enables the system to accurately measure the distance information in three-dimensional space, providing a reliable data basis for subsequent carcass length measurement. The synchronous operation of the camera and the depth camera ensures the real-time acquisition of high-quality RGB and depth map data, avoiding data alignment problems.

[0173] S52, data processing module

[0174] The trained YOLOv8Pose model is used to analyze and process the collected image data in real time. The model can efficiently detect and predict the key point positions of the carcass, and align and fuse the RGB image with the depth image to ensure the accurate position of each key point in three-dimensional space. The distance between key points is calculated using the Euclidean distance algorithm to accurately measure the carcass length. After detecting the key points, the data processing module immediately calculates and records the carcass length to ensure real-time and efficient measurement.

[0175] S53, result display module

[0176] The measurement results are displayed in real time, including drawing key points, connecting lines between key points, and marking the distances between key points. The display content is clear and intuitive, which is convenient for operators to monitor and check the measurement results in real time. At the same time, the analyzed images are stored and visual reports are generated. Key points and their connecting lines are marked in the image, and the distances are marked to provide detailed measurement information. The measurement results are recorded in the form of CSV tables and logs, including detailed data and timestamps for each measurement, which is convenient for subsequent review and data analysis. The system automatically saves and manages measurement data, provides historical data query and analysis functions, and supports breeding feedback and scientific research;

[0177] S54, Data Recording and Storage

[0178] The measurement results are not only displayed in real time, but also stored as visual images, in which key points, key point connections, and distances are clearly drawn. In addition, the measurement results are recorded in CSV tables and logs, including detailed data and timestamps for each measurement, to facilitate subsequent review and data analysis;

[0179] S55, review and analysis function

[0180] The system provides review and analysis functions, can automatically save and manage measurement data, and supports query and analysis of historical data. Operators can view historical measurement data through the system and generate relevant reports to support breeding feedback and scientific research;

[0181] Through the above steps, the present invention realizes the intelligent real-time measurement of carcass length in a complex slaughter line environment, and has the advantages of high precision, real-time and high efficiency. The system not only significantly improves the production efficiency of the slaughter line and reduces labor costs, but also provides scientific feedback data for breeding, and has broad application prospects and economic benefits.

[0182] Furthermore, the S6 body length prediction algorithm uses Euclidean distance to calculate the carcass length. The specific steps are as follows:

[0183] S61, fusion of depth map and RGB map

[0184] The depth camera is used to collect RGB images and depth images at the same time to ensure the synchronization and accuracy of the two. Through image alignment technology, the detection points in the RGB image are accurately matched with the corresponding points in the depth image to ensure that the position of each key point in the three-dimensional space is consistent.

[0185] S62, video stream alignment

[0186] Align the RGB detection points in the video stream with the depth map. Detect the key point positions in the RGB image in real time in the video stream, and obtain the depth information of these key points in 3D space through the corresponding depth image. The alignment process ensures accurate matching of key points in different image sources.

[0187] S63, Euclidean distance calculation

[0188] The Euclidean distance between the key points of the carcass is calculated using the data in the depth map. The specific calculation method is:

[0189] Select multiple key points of the carcass, calculate the three-dimensional coordinates of each key point, and obtain its position in three-dimensional space.

[0190] Calculate the Euclidean distance between adjacent key points, the formula is:

[0191]

[0192] The total length of the carcass can be obtained by adding up the Euclidean distances between all adjacent key points.

[0193] S64, Data Recording and Analysis

[0194] The carcass length data of each measurement is recorded in real time and stored in CSV form and log format. The record content includes detailed data and timestamp of each measurement, which is convenient for subsequent review and data analysis. The system also provides the function of automatically filtering outliers to ensure the accuracy and reliability of the data.

[0195] Through the above steps, the present invention realizes the accurate measurement of the carcass length of the slaughter line, and has the advantages of high precision and high efficiency. By fusing the data of the RGB image and the depth image, the distance between key points is calculated using the Euclidean distance algorithm, which ensures the accuracy and reliability of the carcass length measurement.

[0196] Furthermore, the S7 time series counting method uses the characteristics of the slaughter line to record the number of frames of pigs detected through the ROI (region of interest). Continuous frames represent the same pig, and disconnected frames identify the next pig. This method filters out outliers in data selection to ensure data accuracy and automatic screening. The specific steps are as follows:

[0197] S71, Region of Interest (ROI) Selection

[0198] The selection of the region of interest is based on the actual application scenario of the slaughter line, and is selected in a stable and stable place on the slaughter line. This can effectively record every pig passing through the area and reduce detection errors caused by movement and vibration. The algorithm determines the best ROI area so that every pig can be accurately recorded.

[0199] S72, frame number recording

[0200] In the selected ROI area, the pigs in each frame are detected through the video stream. The consecutively detected frames represent the same pig. When the frames are disconnected, it marks the beginning of the next pig. The system will record the number of frames for each pig passing through the ROI area to ensure the complete record of each pig.

[0201] S73, outlier filtering

[0202] During the recording process, the system automatically filters out outliers to ensure data accuracy. By setting a reasonable frame number threshold, abnormal frames are filtered out to ensure that the data recording of each pig is accurate. The outlier filtering algorithm will be continuously optimized according to the actual application environment to ensure high data quality.

[0203] S74, Data Recording and Storage

[0204] The system stores the data of each pig detected and recorded, and generates a detailed measurement report. The data storage includes the frame number record of the pig, the timestamp of the pig passing through the ROI area, the measured carcass length and other information. These data will be recorded in the form of CSV tables and logs to facilitate subsequent review and data analysis.

[0205] Through the above steps, the time series counting method of the present invention can effectively utilize the characteristics of the slaughter line, accurately record the number of frames of each pig, filter outliers, ensure the accuracy and automatic screening of data. This method not only improves the accuracy and reliability of data recording, but also supports detailed data analysis and breeding research, and has broad application prospects and economic benefits.

[0206] Another object of the present invention is to provide a system for applying the deep learning-based intelligent real-time measurement method for slaughter line carcass length, comprising:

[0207] Image acquisition module: used to use the depth camera to collect three-dimensional data with depth map in real time to ensure the comprehensiveness and accuracy of data acquisition;

[0208] Data processing module: used to use the trained YOLOv8Pose model to analyze image data, predict key points of body length, and measure carcass length. After detecting the key points, the data processing module immediately calculates and records the carcass length, and calculates the distance between key points through the Euclidean distance algorithm to achieve accurate measurement;

[0209] Result display module: used to display the measurement results in real time, including drawing key points, connecting lines between key points, and marking the distance between key points. This module also stores the analyzed images and generates visual reports. The measurement results are recorded in CSV tables and logs to facilitate subsequent review and data analysis;

[0210] Data recording and storage module: used to store the measurement results as visual images, in which key points, key point connections, and distances are clearly drawn. In addition, the measurement results are recorded in CSV tables and logs, including detailed data and timestamps for each measurement, to facilitate subsequent review and data analysis;

[0211] Model visualization module: used to apply visualization algorithms to visualize the model to reveal the decision-making process of the model, ensure the transparency and explainability of the decision, and identify the key facial features that the model relies on through visual analysis.

[0212] Another object of the present invention is to provide a computer device, which includes a memory and a processor, wherein the memory stores a computer program, and when the computer program is executed by the processor, the processor executes the steps of the intelligent real-time measurement method for slaughter line carcass length based on deep learning.

[0213] Another object of the present invention is to provide a computer-readable storage medium storing a computer program. When the computer program is executed by a processor, the processor executes the steps of the intelligent real-time measurement method for slaughter line carcass length based on deep learning.

[0214] Another object of the present invention is to provide an information data processing terminal, which is used to implement the deep learning-based intelligent real-time measurement system for slaughter line carcass length.

[0215] In order to prove the creativity and technical value of the technical solution of the present invention, this section provides application examples of the claimed technical solution on specific products or related technologies.

[0216] Application example: Application case of Muyuan Neixiang slaughterhouse

[0217] Implementation Background

[0218] Muyuan Neixiang Slaughterhouse is a large-scale domestic slaughterhouse with advanced slaughtering production lines. In order to improve production efficiency and measurement accuracy, Muyuan Neixiang Slaughterhouse introduced the deep learning-based intelligent real-time measurement system for slaughtering line carcass length of the present invention.

[0219] System deployment and initialization

[0220] Install depth cameras in key areas: Install depth cameras in key areas of the slaughter line in the slaughterhouse to ensure that the image and depth data of the pig carcasses can be fully captured.

[0221] During the initial deployment, the depth camera is used to collect images and depth data of the pig carcass, and the image is processed and annotated using the method of the present invention to establish an initial data set. This process ensures that the system can adapt to various environments and provides a basis for subsequent real-time measurements.

[0222] Creation of a proprietary dataset: 2,431 carcass images were collected at the Muyuan Neixiang slaughterhouse, covering different environments and lighting conditions, to create a proprietary dataset. These images were preprocessed and data enhanced to ensure the diversity and comprehensiveness of the dataset, providing a solid foundation for subsequent model training.

[0223] Body length key point labeling: Labelme tool is used to label the collected images, and two points of the carcass length, namely the front edge of the pubic symphysis and the front edge of the first cervical vertebra, are selected as key points for labeling. This labeling method improves the prediction accuracy of the model.

[0224] Data enhancement and model training: Apply data enhancement techniques such as color transformation and exposure adjustment to the labeled data set to generate more training samples. Select the YOLOv8Pose model for training and set appropriate hyperparameters such as epochs and batch size to ensure the efficiency and accuracy of the model.

[0225] Real-time monitoring and measurement

[0226] Real-time image and depth data capture: The system captures the image and depth data of the pig carcass in real time during the operation of the slaughter line, and analyzes it through the intelligent measurement system of the present invention to achieve instant measurement and recording of the carcass length.

[0227] The system can efficiently and accurately measure the length of each pig carcass, provide real-time data to support production management, and ensure the consistency and reliability of product quality.

[0228] Construction of real-time measurement system: The system consists of image acquisition module, data processing module and result display module. The image acquisition module uses a depth camera to collect three-dimensional data in real time. The data processing module uses the trained YOLOv8Pose model to analyze images and measure carcass length. The result display module displays the measurement results in real time and stores the analyzed images. The measurement results are recorded in CSV tables and LOG logs.

[0229] Body length prediction and time series counting: Euclidean distance is used to calculate carcass length, and the detection points in the RGB image and the depth image are aligned to accurately measure the carcass length. The time series counting method is used to record the number of pig frames detected by ROI (region of interest) to ensure data accuracy and automatic screening.

[0230] Data analysis and feedback breeding

[0231] Data recording and analysis: Use the measurement data collected by the system for in-depth analysis to provide breeders with detailed feedback information to help optimize breeding strategies and improve breeding efficiency and effectiveness.

[0232] The system is able to record and analyze the data of each measurement. Through data-driven decision support, breeders can better understand the key factors in the breeding process and promote the development of scientific breeding.

[0233] Optimize breeding plans: This data analysis capability significantly improves the scientific nature and efficiency of breeding work, improves and perfects breeding plans, and improves the overall breeding level.

[0234] Product traceability and carcass grading basis

[0235] Detailed measurement records and data storage: The system provides detailed measurement records and data storage functions, providing a scientific basis for carcass grading. Through accurate length measurement, combined with other indicators, the system can grade carcasses more accurately, improving the added value and market recognition of products.

[0236] Full process traceability function: The system's traceability function ensures that the entire process from the pigs entering the slaughtering line to the final product is well documented, enhancing the transparency of product quality and consumer trust.

[0237] Implementation Effect

[0238] Improve production efficiency: The system can quickly output measurement results in the high-frequency production environment of the slaughtering line, avoiding production stagnation caused by measurement delays and significantly improving production efficiency.

[0239] Reduce labor costs: Through full automation and intelligent measurement, the need for manual operation is reduced, reducing labor costs. The number of manual measurement personnel at Muyuan Neixiang Slaughterhouse has been reduced by 30%, significantly saving human resource costs.

[0240] Improve measurement accuracy: The system provides high-precision measurement results, and the measurement error is controlled within ±0.5 cm. Compared with traditional manual measurement methods, the accuracy is improved by more than 50%.

[0241] Data management and visualization: The system can record and analyze measurement data in real time and generate detailed measurement reports. The measurement results are recorded in CSV tables and LOG logs to facilitate subsequent review and data analysis. Muyuan Neixiang slaughterhouse uses this data to optimize carcass grading and breeding strategies, improving production management efficiency.

[0242] Enhanced market competitiveness: By introducing the intelligent measurement system of the present invention, Muyuan Neixiang Slaughterhouse has significantly improved its technical level and production capacity, enhanced its market competitiveness, taken a leading position in the market, and obtained more recognition and orders from customers.

[0243] In summary, through the application and implementation in Muyuan Neixiang slaughterhouse, the deep learning-based intelligent real-time measurement system for slaughter line carcass length of the present invention has demonstrated its significant technical effects and commercial value, and proved its feasibility and superiority in the actual production environment. The system not only improves production efficiency and measurement accuracy, reduces production costs, but also promotes the development of the slaughtering industry towards automation and intelligence, which has important technical significance and application prospects.

[0244] The embodiments of the present invention have achieved some positive effects during the development or use process, and indeed have great advantages over the prior art. The following content is described in conjunction with data, charts, etc. of the test process.

[0245] Significant improvement in recognition accuracy: In the tests conducted, the system's carcass length key point prediction model performed well in terms of accuracy. Figure 7 (A) shows the confusion matrix of the system, indicating that the classification accuracy reaches 1.00 on the test dataset, and (B) shows the F1Confidence curve, showing the performance of the system at different confidence levels.

[0246] Efficiency and cost savings: Theoretically, by implementing this system, the labor efficiency of the pilot slaughterhouse has been significantly improved, and labor costs have been reduced by about 30%. Due to the system's ability to identify diseases early and more optimized production management, overall operating costs have been reduced by about 20%. The system's automated and intelligent measurement reduces reliance on manual measurement and improves overall production efficiency and accuracy.

[0247] Real-time monitoring and data analysis effects: The data collected and analyzed during the experiment showed that the system can effectively monitor the real-time data of pig carcasses and detect abnormal conditions in a timely manner. Through real-time data analysis, the system can provide scientific decision-making support and optimize production processes and management strategies. These data not only help companies improve production efficiency, but also support scientific breeding decisions.

[0248] Product traceability and carcass grading basis: The system provides detailed measurement records and data storage functions, providing a scientific basis for carcass grading. Through accurate length measurement, combined with other indicators, the system can more accurately grade carcasses, improve product added value and market recognition. At the same time, the system's traceability function ensures that the entire process from the pig entering the slaughter line to the final product is well documented, enhancing the transparency of product quality and consumer trust.

[0249] Accuracy performance analysis:

[0250] The effect of the model: Figure 7 (A) and (B) show the performance of the system in model training. The confusion matrix shows that the classification accuracy of the system on the test dataset is 1.00, and the F1Confidence curve shows that the system can maintain efficient measurement performance at different confidence levels.

[0251] Fitting analysis of body length prediction: The following figure shows the test set correlation analysis and ranking correlation analysis of carcass length measurement. It can be seen from the chart that the carcass length predicted by the system has a high correlation with the true value. The left figure shows the fitting curve between the predicted value and the true value, and the fitting degree R 2 is 0.8045. The right figure shows the ranking correlation analysis and the fit R 2 The correlation coefficient r is 0.8135, and the correlation coefficient r reaches 0.90. These results show that the prediction effect of this system in practical application is excellent and can accurately reflect the actual situation of carcass length.

[0252] Through the above tests and analysis, the carcass length intelligent measurement system of the present invention is not only technically innovative and practical, but also shows great potential and value in practical applications. The high precision, high efficiency and high reliability of the system make it have a wide range of application prospects in slaughterhouses, significantly improving production efficiency and product quality, while providing scientific breeding decision support and accurate carcass grading basis for enterprises.

[0253] An embodiment of the present invention provides an intelligent real-time measurement method for slaughter line carcass length based on deep learning. Its core innovation lies in creating a proprietary data set containing 2431 slaughterhouse carcass images, combining the YOLOv8Pose model for deep learning training, and realizing accurate positioning of key points in the carcass image (the front edge of the pubic symphysis and the front edge of the first cervical vertebra), and calculating the carcass length using Euclidean distance. This method improves the generalization ability of the model through data enhancement technology, builds a real-time measurement system, integrates image acquisition, data processing and result display modules, and ensures real-time and accuracy in the slaughter line environment. In addition, the system can be extended to measure other distance traits and area traits, supports a variety of convenient application platforms (such as Android applications and applets), and meets measurement needs in different scenarios.

[0254] 1. Model level: Use "deep learning model" or "artificial intelligence model" as a superordinate concept to cover more types of models.

[0255] 2. Measurement target: Expand to all measurable distance traits and area traits, not just body length.

[0256] 3. Application areas: Expanded from software applications to various convenient application platforms such as Android applications and mini-program applications.

[0257] It should be noted that the embodiments of the present invention can be implemented by hardware, software, or a combination of software and hardware. The hardware part can be implemented using dedicated logic; the software part can be stored in a memory and executed by an appropriate instruction execution system, such as a microprocessor or dedicated design hardware. A person of ordinary skill in the art will appreciate that the above-mentioned devices and methods can be implemented using computer executable instructions and / or contained in a processor control code, such as a carrier medium such as a disk, CD or DVDROM, a programmable memory such as a read-only memory (firmware), or a data carrier such as an optical or electronic signal carrier. Such code is provided on the carrier medium. The device and its modules of the present invention can be implemented by hardware circuits such as very large-scale integrated circuits or gate arrays, semiconductors such as logic chips, transistors, etc., or programmable hardware devices such as field programmable gate arrays, programmable logic devices, etc., can also be implemented by software executed by various types of processors, and can also be implemented by a combination of the above-mentioned hardware circuits and software, such as firmware.

[0258] The above description is only a specific implementation mode of the present invention, but the protection scope of the present invention is not limited thereto. Any modifications, equivalent substitutions and improvements made by any technician familiar with the technical field within the technical scope disclosed by the present invention and within the spirit and principle of the present invention should be covered by the protection scope of the present invention.

Claims

1. A method for intelligent real-time measurement of carcass length on a slaughter line based on deep learning, characterized in that: The proprietary data sets covering different enhancement conditions are trained using the YOLO series or other artificial intelligence models to achieve automatic recognition and accurate measurement of various measurable distance or area traits including carcass length. This method can be used on multiple platforms such as software applications, Android applications, and mini-programs. At the same time, the three-dimensional data collected by the depth camera and the Euclidean distance calculation method between key points are used to ensure the accuracy and real-time nature of the measurement.

2. The method for intelligent real-time measurement of carcass length in a slaughter line based on deep learning according to claim 1, characterized in that: The following steps are involved: S1: RGB images of carcasses split in half under natural light were collected at Muyuan’s slaughterhouse in Neixiang, Henan. After extracting frames, they were optimized and deleted, and data enhancement methods such as color transformation and exposure adjustment were used to finally obtain 2,431 images. The collection process was carried out on a slaughtering line at room temperature, covering carcasses of different sizes to ensure the diversity and comprehensiveness of the data. S2: During the image annotation process, only the inner area of ​​the left fan was annotated to prevent the increase of key point prediction noise due to the detection area being too large, and two key points of the carcass straight length were annotated, the front edge of the pubic symphysis and the front edge of the first cervical vertebra; S3: Increase the diversity of the dataset through data augmentation techniques, including hue, saturation, brightness adjustment intensity, as well as translation, scaling, and mosaic data augmentation; S4: Select and apply the YOLOv8Pose model to train the deep learning model and optimize hyperparameters such as epochs and batch size to ensure the efficiency and accuracy of the model; S5: Build a real-time measurement system, including an image acquisition module, a data processing module, and a result display module. Use a depth camera to collect three-dimensional data with a depth map. The data processing module uses the trained YOLOv8Pose model to analyze the image and measure the carcass length. The result display module displays the measurement results in real time and stores the analyzed images. S6: The carcass length is calculated using the Euclidean distance. By fusing the depth map and the RGB map, the RGB detection points in the video stream are aligned to the corresponding points in the depth map, and the Euclidean distance between the key points of the carcass is calculated using the data in the depth map. S7: The number of pig frames detected by ROI was recorded using the time series counting method to ensure data accuracy and automatic screening.

3. The method for intelligent real-time measurement of carcass length in a slaughter line based on deep learning as claimed in claim 2, characterized in that: In S1, a total of 160 pig images were collected for a total of 15 minutes and 20 seconds. The slaughter line moved forward at a uniform speed. The algorithm extracted the video data of the first 9 minutes and 30 seconds every 2s, extracted 287 images, and selected 221 images. The image video stream after 9 minutes and 30 seconds was used as a test set for inference fitting.

4. The method for intelligent real-time measurement of carcass length in a slaughter line based on deep learning as claimed in claim 2, characterized in that: In S2, the Labelme tool is used for labeling. All detection boxes and key point labels are labeled by one person. The labeling results are saved in Json format and then converted into YOLO format by the algorithm.

5. The method for intelligent real-time measurement of carcass length in a slaughter line based on deep learning as claimed in claim 2, characterized in that: In S3, the data enhancement technology includes using OpenCV to perform color transformation, exposure adjustment, saturation adjustment, brightness change, and simulate image noise and blur on each training set image with key points marked.

6. The method for intelligent real-time measurement of carcass length in a slaughter line based on deep learning as claimed in claim 2, characterized in that: The S4 divides the labeled data into a training set and a validation set in a ratio of 10:1, with the first 100 data as the training set and the validation set, and the last 60 data as the test set. The YOLOv8 model is trained using the ultralytics library, the training cycle is set to 500 epochs, and data enhancement technology is used to improve the robustness of the model.

7. An intelligent real-time measurement system for carcass length in slaughter line based on deep learning, characterized in that: include: Image acquisition module: used to use the depth camera to collect three-dimensional data with depth map in real time to ensure the comprehensiveness and accuracy of data acquisition; Data processing module: used to analyze image data using the trained YOLOv8Pose model, predict body length key points, measure carcass length, and calculate the distance between key points using the Euclidean distance algorithm to achieve accurate measurement; Result display module: used to display the measurement results in real time, including drawing key points, connecting lines between key points, and marking the distance between key points. It also stores the analyzed images and generates visual reports. The measurement results are recorded in CSV tables and logs to facilitate subsequent review and data analysis. Data recording and storage module: used to store the measurement results as visual images, in which key points and key point connections are clearly drawn and distances are marked. The recorded content includes detailed data and timestamps for each measurement.

8. The deep learning-based intelligent real-time measurement system for slaughter line carcass length according to claim 7, characterized in that: The image acquisition module further comprises: Depth camera: used to synchronously capture RGB images and depth images to ensure the synchronization and accuracy of the two. Through image alignment technology, the detection points in the RGB image are accurately matched with the corresponding points in the depth image to ensure that the position of each key point in the three-dimensional space is consistent.

9. The deep learning-based intelligent real-time measurement system for slaughter line carcass length according to claim 8, characterized in that: The data processing module further comprises: Data processing unit: used to analyze the collected image data in real time, use the trained YOLOv8Pose model to efficiently detect and predict the key point positions of the carcass, align and fuse the RGB image with the depth image to ensure the accurate position of each key point in the three-dimensional space, calculate the distance between key points through the Euclidean distance algorithm, and achieve accurate measurement of the carcass length.

10. An information data processing terminal, which is used to implement the deep learning-based intelligent real-time measurement system for carcass length in a slaughter line as described in claim 8.

Citation Information

Patent Citations

  • Pig half-carcass intelligent segmentation method

    CN114342986A

  • Pig body size measurement method based on deep learning multi-parameter fusion

    CN116295022A

  • Livestock body size measuring method, electronic equipment and storage medium

    CN116763295A

  • Slaughter house intelligent monitoring method and system based on video AI analysis

    CN117853998A

  • Livestock body size measuring method based on depth camera and instance segmentation algorithm

    CN118015062A

Cited By

  • Pig carcass size measurement method based on image recognition

    CN119399255A