Machine vision-based Euler sheep body condition score automatic identification system and method

Abstract

The invention relates to the technical field of intelligent livestock breeding, and discloses an Euler sheep body condition score automatic identification system and method based on machine vision. Comprising an automatic identification system, and the automatic identification system comprises a region division module, an image acquisition module, an image preprocessing module, a feature extraction module, a multi-dimensional body condition scoring model and a data output and decision support module. Through the arrangement of the automatic identification system, non-contact automatic acquisition and analysis of Euler sheep body characteristics can be realized through integration of a computer vision technology, a deep learning algorithm and an image processing technology, and a multi-dimensional body condition scoring model is constructed in combination with physiological characteristics and growth stages of the Euler sheep. The method is used for accurately evaluating the fat condition, the health state and the fattening potential of the Tibetan sheep. The system can be applied to large-scale cultivation management, precise nutrition regulation and control and efficient fattening strategy formulation of Tibetan sheep in plateau pasturing areas, and provides intelligent technical support for improving Tibetan sheep cultivation benefits and meat quality.

Description

Technical Field

[0001] The present invention relates to the field of intelligent animal husbandry technology, and more specifically to a machine vision-based automatic recognition system and method for body condition scoring of Euler sheep. Background Art

[0002] Maqu County is a pure pastoral county inhabited mainly by Tibetans. It is one of the important animal husbandry production bases in Gansu Province and the origin of four excellent livestock breeds: Hequ horses, Awancang yaks, Oura sheep and Hequ Tibetan mastiffs. Oura sheep belongs to a local group of Tibetan sheep and is named after the Oura Mountains on the border between Maqu County, Gansu Province and Henan County, Qinghai Province. Oura-type Tibetan sheep are known for their large size, strong resistance to adversity and good meat performance. They are mainly distributed in Oura Township, Maqu County. Compared with the meat of plateau-type Tibetan sheep, the meat of Oura-type Tibetan sheep is of superior quality, fragrant and beautiful taste, rich in nutrition, and is semi-game. It is especially characterized by high energy, high protein, high mineral content, low fat and low cholesterol, and is richer in amino acids and heme iron. It has health values such as anti-cancer, anti-aging, iron supplementation and physical recovery, and is known as "plateau green food"; In the current breeding model for Oula sheep, herders rely mainly on manual experience to visually assess the physical condition of Oula sheep, judging their nutritional status and fattening effect by observing characteristics such as the sheep's body outline, back fat thickness, and rib exposure. However, this manual scoring method has significant flaws: Highly subjective: Different breeders have different experiences, which results in different tastes of different Oula sheep after being raised; Inefficiency: Manual observation of each sheep is time-consuming and labor-intensive, and cannot meet the needs of rapid screening of large flocks; Data loss: Lack of quantitative recording and analysis of body characteristics makes it difficult to establish a dynamic growth model to optimize fattening strategies; Therefore, there is an urgent need for an automatic recognition system and method for the body condition score of Oura sheep based on machine vision to solve the technical problems raised above. Summary of the Invention

[0003] In order to overcome the above-mentioned defects of the prior art, the implementation regulations of the present invention provide a system and method for automatic identification of body condition scores of Oura sheep based on machine vision to solve the technical problems raised in the background technology.

[0004] To achieve the above-mentioned object, the present invention provides the following technical solution: a system and method for automatically identifying the body condition score of Euler sheep based on machine vision, comprising an automatic identification system, the automatic identification system including a region division module, an image acquisition module, an image preprocessing module, a feature extraction module, a multidimensional body condition scoring model, and a data output and decision support module, wherein a feeding management system is provided inside the region division module; The area division module divides the breeding area into n groups of sub-breeding areas (Z1-Z n ); the image acquisition module is used to collect the corresponding data of the side view, top view, rear view and sheep movement posture of the n groups of Oula sheep in the n groups of sub-breeding areas in real time; the image preprocessing module is used to process the corresponding data in the image acquisition module, and extract the corresponding back fat thickness estimation value, rib exposure index, waist-to-hip ratio, shoulder peak height and hair coverage density generated by the Oula sheep in each group of sub-breeding areas; the multidimensional body condition scoring model is used to integrate the body characteristic parameters, preset physiological stage weight factors and historical growth data, and output fatness score, health status index and fattening potential prediction value; the decision support module is used to generate a visual scoring report, and work in conjunction with the feeding management system to output precise nutritional regulation suggestions.

[0005] In a preferred embodiment, the feeding management system includes a formula database and a multi-channel feeding device, wherein the feeding management system stores the formulas of each sub-breeding area (Z1-Z n ) corresponding differentiated feed formula set, the differentiated feed formula set including at least the nutrient element content ratio, the particle size distribution coefficient and the additive concentration threshold; the multi-channel feeding device is connected to the formula database signal, the multi-channel feeding device includes m independently controlled feeding units, the m groups of feeding units are sequentially arranged in the n groups of sub-breeding areas, and the differentiated feeding operation of the designated sub-breeding area is performed according to the differentiated feed formula set provided by the formula database.

[0006] In a preferred embodiment, the image acquisition module includes a multi-view HD camera group, an infrared sensor, and an ambient light compensation device; the multi-view HD camera group includes three cameras and is respectively located at the top, side, and rear of the sheep passage, covering multi-angle images of the Euler sheep when walking, standing, and feeding; The infrared sensor is used to trigger the camera to automatically take pictures when the sheep enters the preset area and simultaneously record the ambient light intensity; The ambient light compensation device eliminates the effects of strong ultraviolet rays on the plateau, rainy weather, and day-night lighting differences on image quality by dynamically adjusting camera exposure parameters and fill light intensity.

[0007] In a preferred embodiment, the image preprocessing unit module is communicatively connected to the image acquisition module, and the image preprocessing module includes a data denoising unit communicatively connected thereto, configured to receive the original image data transmitted by the image acquisition module, and perform a noise suppression operation based on digital image processing technology on the original image data to generate target image data after noise reduction.

[0008] In a preferred embodiment, the data denoising unit uses a Wiener filter to perform denoising on the original image data, wherein the data denoising unit includes a signal receiving module, a spectrum analysis module and a filtering module; The signal receiving module is used to obtain the noisy image signal input through the transmission channel and convert the image signal into frequency domain data; Spectrum analysis module, configured to obtain noise power spectrum density through pre-calibration or dynamic analysis ,in Represents the spatial frequency component of the image; Frequency domain data based on noisy image signals , generate the original signal power spectrum density through smoothing or historical data statistics ; Filter module: According to the formula , Construct a two-dimensional frequency domain filter; The noisy image spectrum and filter kernel Perform point-by-point multiplication to generate denoised real-time image data.

[0009] In a preferred embodiment, the processing steps of the image pre-processing unit include: The improved U-Net model is used to perform background segmentation on the denoised image data to distinguish between grass, rocks and sheep outlines; The segmented sheep body image is processed for illumination equalization based on the Retinex algorithm to enhance the back fat thickness and texture details of the rib area. The scapula, lumbar spinous process and ischial tuberosity of the Euler sheep were located through the key point detection algorithm, and the backfat measurement line and waist-to-hip ratio calculation area were delineated.

[0010] In a preferred embodiment, the feature extraction module includes: using a convolutional neural network with ResNet-50 as the backbone network and embedding a channel attention mechanism to enhance sensitivity to small changes in backfat thickness; The feature extraction module also includes: calculating and predicting backfat thickness through a regression model, calculating the rib exposure index based on morphological operations, and quantifying waist-to-hip ratio and acromion height in combination with a geometric transformation algorithm.

[0011] In a preferred embodiment, the multidimensional body condition scoring model is established by the following steps: S1. Establishing a baseline curve of body characteristics of Oula sheep at different physiological stages based on historical data; S2. Use the random forest algorithm to integrate body shape characteristic parameters, ambient temperature, altitude and feeding records to output fatness score; S3. Analyze the changing trends of continuously collected body shape characteristics through the LSTM network to predict the fattening weight gain potential and health risk level in the next 30 days.

[0012] In a preferred embodiment, the data output and decision support module summarizes the best differentiated feed formula set based on the fatness score and generates the best feeding plan, which includes feed ratio and ambient temperature. The data output and decision support module automatically marks low-scoring sheep and pushes veterinary inspection reminders.

[0013] The automatic recognition method of the body condition score of the Euler sheep based on machine vision is applied to the automatic recognition system of the body condition score of the Euler sheep based on machine vision, and includes the following steps: Step 1: Divide the breeding area into n groups of sub-breeding areas through the regional division module, and use the feeding management system to carry out breeding with corresponding differentiated feed formulas for each sub-breeding area; Step 2: The image acquisition module obtains the real-time dynamic image of the Euler sheep; Step 3: The image preprocessing module performs denoising on the image data, and performs background segmentation, illumination compensation, and key anatomical location. Step 4: The feature extraction module extracts quantitative features of backfat thickness, rib exposure index, waist-to-hip ratio, and acromion height; Step 5: Combine physiological stages and environmental parameters to output fatness grade and fattening recommendations through a multidimensional scoring model; Step 6: Link fatness level with feeding management system and health monitoring platform to achieve accurate breeding decision-making.

[0014] Technical effects and advantages of the present invention: The present invention incorporates an automatic recognition system that facilitates contactless automated collection and analysis of the body characteristics of Tibetan sheep by integrating computer vision, deep learning algorithms, and image processing techniques. This system, combined with their physiological characteristics and growth stages, constructs a multidimensional body condition scoring model for accurately assessing their fatness, health, and fattening potential. This system can be applied to large-scale Tibetan sheep farming management, precise nutritional regulation, and efficient fattening strategy development in highland pastoral areas, providing intelligent technical support for improving Tibetan sheep farming efficiency and meat quality. This invention completely replaces manual visual scoring through multi-view image acquisition and deep learning-driven feature extraction, significantly reducing subjective errors and solving the pain points of low efficiency and data loss in fatness assessment under traditional breeding models. The present invention integrates body parameters, feeding records and environmental data extracted by machine vision through a random forest algorithm to comprehensively determine the fatness level, with a model accuracy rate of 92%. Through the real-time linkage of fatness scoring and feeding management system, differentiated daily diet ratios are achieved, which increases feed conversion rate by 15%-20% and shortens fattening cycle by 10%-15%. BRIEF DESCRIPTION OF THE DRAWINGS

[0015] Figure 1 This is the overall flow chart of the automatic identification system of the present invention.

[0016] Figure 2 This is a schematic diagram of the overall process of the method for automatically identifying the body condition score of Oura sheep of the present invention. DETAILED DESCRIPTION

[0017] The technical solutions of the present invention will be clearly and completely described below in conjunction with the accompanying drawings of the present invention. In addition, the forms of the various structures described in the following embodiments are merely examples. The automatic identification system and method for body condition scoring of Oura sheep based on machine vision involved in the present invention are not limited to the various structures described in the following embodiments. All other embodiments obtained by ordinary technicians in this field without making creative work are within the scope of protection of the present invention.

[0018] The present invention provides a system and method for automatically identifying the body condition score of an Oula sheep based on machine vision, including an automatic identification system, wherein the automatic identification system includes a region division module, an image acquisition module, an image preprocessing module, a feature extraction module, a multidimensional body condition scoring model, and a data output and decision support module, wherein a feeding management system is provided inside the region division module; The area division module divides the breeding area into n groups of sub-breeding areas (Z1-Z n ); the image acquisition module is used to collect the corresponding data of the side view, top view, rear view and sheep movement posture of the n groups of Oula sheep in the n groups of sub-breeding areas in real time; the image preprocessing module is used to process the corresponding data in the image acquisition module, and extract the corresponding back fat thickness estimation value, rib exposure index, waist-to-hip ratio, shoulder peak height and hair coverage density generated by the Oula sheep in each group of sub-breeding areas; the multidimensional body condition scoring model is used to integrate the body characteristic parameters, preset physiological stage weight factors and historical growth data, and output fatness score, health status index and fattening potential prediction value; the decision support module is used to generate a visual scoring report, and work in conjunction with the feeding management system to output precise nutritional regulation suggestions.

[0019] Reference Figure 1-2As shown, the present invention provides an automatic recognition system and method for body condition scoring of Euler sheep based on machine vision, wherein the feeding management system includes a formula database and a multi-channel feeding device, wherein the feeding management system stores information related to each sub-breeding area (Z1-Z n ) a corresponding differentiated feed formula set, the differentiated feed formula set including at least a nutrient element content ratio, a particle size distribution coefficient, and an additive concentration threshold; the multi-channel feeding device is signal-connected to the formula database, the multi-channel feeding device includes N independently controlled feeding units, each group of feeding units is sequentially arranged in n groups of sub-breeding areas, and performs differentiated feeding operations in designated sub-breeding areas according to the differentiated feed formula provided by the formula database; The image acquisition module includes a multi-view HD camera group, an infrared sensor, and an ambient light compensation device. The multi-view HD camera group includes three cameras located at the top, side, and rear of the sheep passage, covering multi-angle images of the sheep walking, standing, and feeding. The infrared sensor is used to trigger the camera to automatically take pictures when the sheep enters the preset area and simultaneously record the ambient light intensity; The ambient light compensation device eliminates the effects of strong ultraviolet rays on the plateau, rainy weather, and day-night lighting differences on image quality by dynamically adjusting camera exposure parameters and fill light intensity.

[0020] The image preprocessing unit module is communicatively connected to the image acquisition module, and the image preprocessing module includes a data denoising unit communicatively connected thereto, configured to receive the original image data transmitted by the image acquisition module and perform a noise suppression operation based on digital image processing technology on the original image data to generate target image data after noise reduction.

[0021] The data denoising unit uses a Wiener filter to perform denoising on the original image data, wherein the data denoising unit includes a signal receiving module, a spectrum analysis module and a filtering module; The signal receiving module is used to obtain the noisy image signal input through the transmission channel and convert the image signal into frequency domain data; Spectrum analysis module, configured to obtain noise power spectrum density through pre-calibration or dynamic analysis ,in Represents the spatial frequency component of the image; Frequency domain data based on noisy image signals , generate the original signal power spectrum density through smoothing or historical data statistics ; Filter module: According to the formula Construct a two-dimensional frequency domain filter; The noisy image spectrum and filter kernel Perform point-by-point multiplication to generate denoised real-time image data.

[0022] The processing steps of the image preprocessing unit include: The improved U-Net model is used to perform background segmentation on the denoised image data to distinguish between grass, rocks and sheep outlines; The segmented sheep body image is processed for illumination equalization based on the Retinex algorithm to enhance the back fat thickness and texture details of the rib area. The scapula, lumbar spinous process and ischial tuberosity of the Euler sheep were located through the key point detection algorithm, and the backfat measurement line and waist-to-hip ratio calculation area were delineated.

[0023] The feature extraction module includes: using a convolutional neural network with ResNet-50 as the backbone network and embedding a channel attention mechanism to enhance sensitivity to small changes in backfat thickness; The feature extraction module also includes: calculating and predicting backfat thickness through a regression model, calculating the rib exposure index based on morphological operations, and quantifying waist-to-hip ratio and acromion height in combination with a geometric transformation algorithm.

[0024] The multidimensional body condition scoring model is established by the following steps: S1. Establishing a baseline curve of body characteristics of Oula sheep at different physiological stages based on historical data; S2. Use the random forest algorithm to integrate body shape characteristic parameters, ambient temperature, altitude and feeding records to output fatness score; S3. Analyze the changing trends of continuously collected body shape characteristics through the LSTM network to predict the fattening weight gain potential and health risk level in the next 30 days.

[0025] The data output and decision support module summarizes the best differentiated feed formula set based on fatness scores and generates the best feeding plan, which includes feed ratio and ambient temperature. The data output and decision support module automatically marks low-scoring sheep and pushes veterinary inspection reminders.

[0026] The automatic recognition method of the body condition score of the Euler sheep based on machine vision is applied to the automatic recognition system of the body condition score of the Euler sheep based on machine vision, and includes the following steps: Step 1: Divide the breeding area into n groups of sub-breeding areas through the regional division module, and use the feeding management system to carry out breeding with corresponding differentiated feed formulas for each sub-breeding area; Step 2: The image acquisition module obtains the real-time dynamic image of the Euler sheep; Step 3: The image preprocessing module performs denoising on the image data, and performs background segmentation, illumination compensation, and key anatomical location. Step 4: The feature extraction module extracts quantitative features of backfat thickness, rib exposure index, waist-to-hip ratio, and acromion height; Step 5: Combine physiological stages and environmental parameters to output fatness grade and fattening recommendations through a multidimensional scoring model; Step 6: Link fatness level with feeding management system and health monitoring platform to achieve accurate breeding decision-making.

[0027] In the embodiment of the present application, the feed formula in the differentiated feed formula set is the same as the composition of the local forage grass, and the proportion of the formula content in each differentiated feed formula set is different.

[0028] The feeding unit is equipped with a weighing sensor and a visual recognition component. The weighing sensor detects the deviation between the actual feeding amount and the target formula. When the deviation value is ≥5%, the visual recognition component is triggered to perform particle size distribution compliance verification, so as to improve the accuracy of the visual scoring report generated in each sub-breeding area. Among them, the different physiological stages of different Euler sheep are: lamb, growing sheep and adult sheep; The fatness score is: 1-5; Grade 1 (emaciated): ribs are obviously exposed, back fat thickness is less than 5mm, and waist-to-hip ratio is abnormally low; Level 2 (slightly thin): ribs are visible but not sunken, back fat thickness is 5-10mm; Level 3 (moderate): Ribs are not visible but palpable, backfat thickness is 10-15 mm, and waist-to-hip ratio is consistent with the physiological stage; Level 4 (fat): backfat thickness 15-20mm, high waist-to-hip ratio; Level 5 (overfat): Back fat thickness > 20mm, waist-to-hip ratio is significantly abnormal.

[0029] The above embodiments can be implemented in whole or in part by software, hardware, firmware or any other combination thereof. Those skilled in the art will appreciate that the units and algorithm steps of the various examples described in conjunction with the embodiments disclosed herein can be implemented by electronic hardware or a combination of computer software and electronic hardware. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. Professional and technical personnel can use different methods to implement the described functions for each specific application, but such implementation should not be considered to be beyond the scope of this application.

[0030] Those skilled in the art will clearly understand that, for the convenience and brevity of description, the specific working processes of the systems, devices and units described above can refer to the corresponding processes in the aforementioned method embodiments and will not be repeated here.

[0031] In the several embodiments provided in this application, it should be understood that the disclosed systems, devices and methods can be implemented in other ways. For example, the device embodiments described above are merely schematic. For example, the division of the units is only a logical function division. There may be other division methods in actual implementation, such as multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be through some interfaces, indirect coupling or communication connection of devices or units, which can be electrical, mechanical or other forms.

[0032] The units described as separate components may or may not be physically separate, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed across multiple network units. Some or all of these units may be selected to achieve the purpose of this embodiment according to actual needs.

[0033] In addition, each functional unit in each embodiment of the present application may be integrated into one processing unit, or each unit may exist physically separately, or two or more units may be integrated into one unit.

[0034] The above description is merely a specific embodiment of the present application, but the scope of protection of the present application is not limited thereto. Any changes or substitutions that can be easily conceived by a person skilled in the art within the technical scope disclosed in this application should be included in the scope of protection of this application. Therefore, the scope of protection of this application should be based on the scope of protection of the claims.

[0035] Finally: The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present invention should be included in the scope of protection of the present invention.

Claims

1. The automatic recognition system for body condition scoring of Euler sheep based on machine vision is characterized by: The automatic recognition system includes an area division module, an image acquisition module, an image preprocessing module, a feature extraction module, a multidimensional body condition scoring model, and a data output and decision support module, wherein a feeding management system is provided inside the area division module; The area division module divides the breeding area into n groups of sub-breeding areas (Z1-Z n ); the image acquisition module is used to collect the corresponding data of the side view, top view, rear view and sheep movement posture of the n groups of Oula sheep in the n groups of sub-breeding areas in real time; the image preprocessing module is used to process the corresponding data in the image acquisition module, and extract the corresponding back fat thickness estimation value, rib exposure index, waist-to-hip ratio, shoulder peak height and hair coverage density generated by the Oula sheep in each group of sub-breeding areas; the multidimensional body condition scoring model is used to integrate the body characteristic parameters, preset physiological stage weight factors and historical growth data, and output fatness score, health status index and fattening potential prediction value; the decision support module is used to generate a visual scoring report, and work in conjunction with the feeding management system to output precise nutritional regulation suggestions.

2. The automatic recognition system for body condition scoring of Euler sheep based on machine vision according to claim 1, characterized in that: The feeding management system includes a formula database and a multi-channel feeding device, wherein the feeding management system stores information related to each sub-breeding area (Z1-Z n ) corresponding differentiated feed formula set, the differentiated feed formula set including at least the nutrient element content ratio, the particle size distribution coefficient and the additive concentration threshold; the multi-channel feeding device is connected to the formula database signal, the multi-channel feeding device includes m independently controlled feeding units, the m groups of feeding units are sequentially arranged in the n groups of sub-breeding areas, and the differentiated feeding operation of the designated sub-breeding area is performed according to the differentiated feed formula set provided by the formula database.

3. The automatic recognition system for body condition scoring of Euler sheep based on machine vision according to claim 1, characterized in that: The image acquisition module includes a multi-view HD camera group, an infrared sensor, and an ambient light compensation device. The multi-view HD camera group includes three cameras located at the top, side, and rear of the sheep passage, covering multi-angle images of the sheep walking, standing, and feeding. The infrared sensor is used to trigger the camera to automatically take pictures when the sheep enters the preset area and simultaneously record the ambient light intensity; The ambient light compensation device eliminates the effects of strong ultraviolet rays on the plateau, rainy weather, and day-night lighting differences on image quality by dynamically adjusting camera exposure parameters and fill light intensity.

4. The automatic recognition system for body condition scoring of Euler sheep based on machine vision according to claim 1, characterized in that: The image preprocessing unit module is communicatively connected to the image acquisition module, and the image preprocessing module includes a data denoising unit communicatively connected thereto, configured to receive the original image data transmitted by the image acquisition module and perform a noise suppression operation based on digital image processing technology on the original image data to generate target image data after noise reduction.

5. The automatic recognition system for body condition scoring of Euler sheep based on machine vision according to claim 4 is characterized in that: The data denoising unit uses a Wiener filter to perform denoising on the original image data, wherein the data denoising unit includes a signal receiving module, a spectrum analysis module and a filtering module; The signal receiving module is used to obtain the noisy image signal input through the transmission channel and convert the image signal into frequency domain data; Spectrum analysis module, configured to obtain noise power spectrum density through pre-calibration or dynamic analysis ,in Represents the spatial frequency component of the image; Frequency domain data based on noisy image signals , generate the original signal power spectrum density through smoothing or historical data statistics ; Filter module: According to the formula ， Construct a two-dimensional frequency domain filter; The noisy image spectrum and filter kernel Perform point-by-point multiplication to generate denoised real-time image data.

6. The automatic recognition system for body condition scoring of Euler sheep based on machine vision according to claim 5, characterized in that: The processing steps of the image preprocessing unit include: The improved U-Net model is used to perform background segmentation on the denoised image data to distinguish between grass, rocks and sheep outlines; The segmented sheep body image is processed for illumination equalization based on the Retinex algorithm to enhance the back fat thickness and texture details of the rib area. The scapula, lumbar spinous process and ischial tuberosity of the Euler sheep were located through the key point detection algorithm, and the backfat measurement line and waist-to-hip ratio calculation area were delineated.

7. The automatic recognition system for body condition scoring of Euler sheep based on machine vision according to claim 1, characterized in that: The feature extraction module includes: using a convolutional neural network with ResNet-50 as the backbone network and embedding a channel attention mechanism to enhance sensitivity to small changes in backfat thickness; The feature extraction module also includes: calculating and predicting backfat thickness through a regression model, calculating the rib exposure index based on morphological operations, and quantifying waist-to-hip ratio and acromion height in combination with a geometric transformation algorithm.

8. The automatic recognition system for body condition scoring of Euler sheep based on machine vision according to claim 1, characterized in that: The multidimensional body condition scoring model is established by the following steps: S1. Establishing a baseline curve of body characteristics of Oula sheep at different physiological stages based on historical data; S2. Use the random forest algorithm to integrate body shape characteristic parameters, ambient temperature, altitude and feeding records to output fatness score; S3. Analyze the changing trends of continuously collected body shape characteristics through the LSTM network to predict the fattening weight gain potential and health risk level in the next 30 days.

9. The automatic recognition system for body condition scoring of Euler sheep based on machine vision according to claim 1, characterized in that: The data output and decision support module summarizes the best differentiated feed formula set based on fatness scores and generates the best feeding plan, which includes feed ratio and ambient temperature. The data output and decision support module automatically marks low-scoring sheep and pushes veterinary inspection reminders.

10. A method for automatically identifying the body condition score of an Oula sheep based on machine vision, applied to the automatic identification system for the body condition score of an Oula sheep based on machine vision according to claims 1 to 9, comprising the following steps: Step 1: Divide the breeding area into n groups of sub-breeding areas through the regional division module, and use the feeding management system to carry out breeding with corresponding differentiated feed formulas for each sub-breeding area; Step 2: The image acquisition module obtains the real-time dynamic image of the Euler sheep; Step 3: The image preprocessing module performs denoising on the image data, and performs background segmentation, illumination compensation, and key anatomical location. Step 4: The feature extraction module extracts quantitative features of backfat thickness, rib exposure index, waist-to-hip ratio, and acromion height; Step 5: Combine physiological stages and environmental parameters to output fatness grade and fattening recommendations through a multidimensional scoring model; Step 6: Link fatness level with feeding management system and health monitoring platform to achieve accurate breeding decision-making.

Images