Chicken body size and weight evaluation system and method based on image recognition

Abstract

The invention relates to the technical field of chicken body and weight evaluation and detection, and provides a chicken body and weight evaluation system and method based on image recognition, and the method comprises the steps: extracting a plurality of chicken images in chicken coop images, and calculating the 2D features and 3D features of a chicken corresponding to each chicken image through the depth information of the chicken coop images, and then generating a 3D model of each chicken by using the 2D features and the 3D features and calculating a weight parameter, and finally judging whether to feed back an evaluation alarm signal according to the number of chicken images with abnormal size parameters and / or weight parameters in the plurality of chicken house images. Therefore, by adopting a mode of combining image recognition and depth information, the position of the image sensor in the chicken house is not restrained any more, the scene adaptability is improved, meanwhile, the volume is corrected by generating the 3D model of each chicken, and the weight parameter is determined based on the conversion relation between the model unit volume and the weight, so that the accuracy of the weight parameter is improved. And the body shape and weight evaluation accuracy of a large number of chickens in the henhouse is improved.

Description

Technical Field

[0001] The present invention relates to the technical field of chicken body shape and weight assessment, and in particular to a chicken body shape and weight assessment system and method based on image recognition. Background Art

[0002] The assessment of chicken body size and weight plays a core role in the breeding industry. Accurate measurement of body size and weight can directly reflect the growth status and nutritional level of chickens, providing a scientific basis for feeding and management. Changes in body size and weight directly reflect the growth rate and health status of chickens, allowing breeders to adjust feed formulas and feeding environments in a timely manner to ensure that chickens can obtain optimal nutritional support at all growth stages. Secondly, body size and weight are closely related to key production indicators such as egg production rate, egg weight, and meat quality of chickens. Through reasonable evaluation of these indicators, breeders can optimize feeding strategies and select more suitable breeds for breeding, thereby improving overall economic benefits. For example, larger chickens may lay more eggs, while chickens with moderate weight have better chicken quality, all of which require accurate evaluation to make judgments. In addition, the assessment of body size and weight can also help with early warning of diseases and breed improvement. Sudden changes in weight may indicate health problems. Timely detection and treatment can reduce disease transmission and reduce economic losses.

[0003] Although there are many animal and plant body shape or weight detection methods on the market, such as CN118038160A proposed a tuna weight estimation method based on deep learning algorithm, which estimates the weight of tuna by converting pixel size into actual distance. In the chicken breeding scenario, the number of chickens in a single chicken house can reach hundreds or even thousands, but the existing methods basically detect single or a small number of animals and plants, and do not have the ability to detect the body shape or weight of a large number of animals and plants at the same time.

[0004] Therefore, how to improve the accuracy of size and weight assessment of a large number of chickens in a chicken house and provide accurate data support for chicken breeding and management is a technical problem that needs to be solved urgently. Summary of the invention

[0005] In order to solve the above-mentioned problems of the prior art, the present invention provides a chicken body shape and weight assessment system and method based on image recognition, which aims to solve the problems of low detection accuracy, heavy computational load and high equipment cost in the animal and plant shape or weight detection methods provided by the prior art when detecting a large number of animals and plants.

[0006] The present invention provides a chicken body shape and weight assessment system based on image recognition, comprising: An image acquisition and processing module, wherein the image acquisition and processing module is configured to acquire a plurality of chicken house images of a target chicken house, identify and segment a plurality of chicken images in each chicken house image, and extract depth information from the chicken house images; A size calculation module, wherein the size calculation module is configured to calculate the size parameters of each chicken in each chicken house image based on the plurality of chicken images obtained by segmentation and the depth information; wherein the size parameters include 2D features and 3D features of each chicken; A weight estimation module, wherein the weight estimation module is configured to generate a 3D model of each chicken based on the 2D features and 3D features of each chicken in each chicken house image, correct the size parameters of each chicken and calculate the weight parameters of each chicken; The result feedback module is configured to determine whether there are chicken images with abnormal size parameters and / or weight parameters exceeding a preset number in each chicken house image collected at each moment, and if so, feedback evaluation alarm information.

[0007] Optionally, the image acquisition and processing module specifically includes: Image acquisition unit; Wherein, the image acquisition unit is configured to acquire images of the chicken house acquired by a plurality of image sensors arranged at different positions of the target chicken house; The chicken house images collected by several image collectors are configured as several images covering different ranges of the target chicken house.

[0008] Optionally, the image acquisition and processing module specifically includes: Image segmentation unit; The image segmentation unit is configured to use a PIDNet model trained based on chicken images to identify and segment chicken images in a plurality of chicken house images of a target chicken house, and obtain a chicken image detection result for each image; The chicken image detection result is specifically: , , …, ; in, represents the chicken image detection result of the Mth chicken house image, The mask representing the Nth chicken detected in the Mth chicken house image.

[0009] Optionally, the image acquisition and processing module specifically includes: Depth estimation unit; The depth estimation unit is configured to extract the depth information of each chicken house image using the SideRT model and store it in the form of a depth map.

[0010] Optionally, the size measurement module specifically includes: 2D feature extraction unit; The 2D feature extraction unit is configured to perform scaling processing on the information extraction target generated by the mask based on each chicken image based on the depth information, and calculate and obtain the 2D features of each chicken image; The 2D features include: the projected area S, the width W of the chicken image, and the perimeter L of the chicken image, and the specific expression is: ; ; ; in, is the area of ​​the mask, is the width of the mask, is the perimeter of the mask, D is the average depth value of the mask location, k is a constant for adjusting the degree of nonlinearity, and b are scaling factors used to adjust the effect of depth on area, and c is an offset used to prevent the area from being scaled too large when the depth value is too small.

[0011] Optionally, the size measurement module specifically includes: 3D feature extraction unit; Wherein, the 3D feature extraction unit is configured to estimate the 3D features of each chicken image according to the depth information of each chicken house image and the projection area S; The 3D features include: volume V, which is specifically expressed as: The 3D features include: volume V, which is specifically expressed as: ; Where V is the estimated volume, represents the average depth value of the mask perimeter of each chicken image, Represents the depth value of the i-th pixel at the corresponding position of the mask of each chicken image.

[0012] Optionally, the weight estimation module specifically includes: A 3D model generation unit, wherein the 3D model generation unit is configured to generate a feature data set based on the 2D features and 3D features of each chicken in each chicken house image, input the feature data set into a neural radiation field model to obtain a feature potential representation, and input the feature potential representation into a CLIP-Mesh model to obtain a 3D model of each chicken; A volume correction unit is configured to correct the 3D features of each chicken into a volume corresponding to the 3D model according to the 3D model of each chicken.

[0013] Optionally, the weight estimation module specifically includes: Weight calculation unit; The weight calculation unit is configured to calculate the weight parameters of each chicken based on the 3D model of each chicken and a preset ratio of weight to unit volume.

[0014] Optionally, the result feedback module specifically includes: An abnormality determination unit, the abnormality determination unit is configured to determine whether the size parameter and / or weight parameter of each chicken is an abnormal value, and if so, determine the chicken as an abnormal target; An evaluation alarm unit is configured to determine whether the number of chickens determined to be abnormal targets in each chicken house image collected at each moment exceeds a preset number, and if so, feed back an evaluation alarm signal.

[0015] The present invention also provides a chicken body shape and weight assessment method based on image recognition, comprising: S1: Acquire several chicken house images of the target chicken house, identify and segment several chicken images in each chicken house image, and extract depth information in the chicken house image; S2: Based on the plurality of chicken images obtained by segmentation and the depth information, the size parameters of each chicken in each chicken house image are calculated; wherein the size parameters include 2D features and 3D features of each chicken; S3: Based on the 2D features and 3D features of each chicken in each chicken house image, a 3D model of each chicken is generated, the size parameters of each chicken are corrected, and the weight parameters of each chicken are calculated; S4: Determine whether there are chicken images with abnormal size parameters and / or weight parameters exceeding a preset number in each chicken house image collected at each moment, and if so, feedback evaluation alarm information.

[0016] The beneficial effects of the present invention are as follows: a chicken body shape and weight assessment system and method based on image recognition is proposed, by extracting several chicken images in the chicken house image, using the depth information of the chicken house image to measure the 2D features and 3D features of the chicken corresponding to each chicken image, and then using the 2D features and 3D features to generate a 3D model of each chicken and calculate the weight parameters, and finally according to the number of chicken images with abnormal size parameters and / or weight parameters in several chicken house images, it is determined whether to feedback an assessment alarm signal. Thus, by combining image recognition with depth information, the position of the image sensor in the chicken house is no longer constrained, and the scene adaptability is improved. At the same time, by generating a 3D model of each chicken to correct the volume, and determining the weight parameters based on the conversion relationship between the unit volume and weight of the model, the accuracy of the body shape and weight assessment of a large number of chickens in the chicken house is improved. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] Figure 1 This is a schematic diagram of the structure of the chicken shape and weight assessment system based on image recognition provided in this embodiment.

[0018] Figure 2 This is a schematic diagram of an example of the placement of image sensors in a chicken house in this embodiment; Figure 3 Schematic diagram of the monitoring range of each image sensor in the chicken house in this embodiment; Figure 4 This is a schematic flow chart of the chicken body shape and weight assessment method based on image recognition provided in this embodiment.

[0019] Reference numerals: 10-image acquisition and processing module; 20-size measurement module; 30-weight estimation module; 40-result feedback module. DETAILED DESCRIPTION

[0020] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0021] Embodiment 1: like Figure 1 As shown, a chicken body shape and weight assessment system based on image recognition includes: an image acquisition and processing module 10, wherein the image acquisition and processing module 10 is configured to obtain a plurality of chicken house images of a target chicken house, identify and segment a plurality of chicken images in each chicken house image, and extract depth information in the chicken house image; a size measurement module 20, wherein the size measurement module 20 is configured to measure the size parameters of each chicken in each chicken house image based on the plurality of chicken images obtained by segmentation and the depth information; wherein the size parameters include 2D features and 3D features of each chicken; a weight estimation module 30, wherein the weight estimation module 30 is configured to generate a 3D model of each chicken based on the 2D features and 3D features of each chicken in each chicken house image, correct the size parameters of each chicken and calculate the weight parameters of each chicken; a result feedback module 40, wherein the result feedback module 40 is configured to determine whether there are chicken images with size parameters and / or abnormal weight parameters exceeding a preset number in each chicken house image acquired at each moment, and if so, feedback assessment alarm information.

[0022] It should be noted that the animal and plant body shape or weight detection method provided by the prior art has the problems of low detection accuracy, large computing load and high equipment cost when detecting a large number of animals and plants. In order to solve the above problems, this embodiment extracts several chicken images from the chicken house image, uses the depth information of the chicken house image to measure the 2D features and 3D features of the chicken corresponding to each chicken image, and then uses the 2D features and 3D features to generate a 3D model of each chicken and calculate the weight parameter, and finally determines whether to feedback the evaluation alarm signal according to the number of chicken images with abnormal size parameters and / or weight parameters in the several chicken house images. Therefore, by combining image recognition with depth information, the position of the image sensor in the chicken house is no longer constrained, and the scene adaptability is improved. At the same time, the volume is corrected by generating a 3D model of each chicken, and the weight parameter is determined based on the conversion relationship between the unit volume and weight of the model, thereby improving the accuracy of the body shape and weight assessment of a large number of chickens in the chicken house.

[0023] In a preferred embodiment, the image acquisition and processing module specifically includes: an image acquisition unit; wherein the image acquisition unit is configured to obtain images of a chicken house acquired by several image sensors arranged at different positions of the target chicken house; wherein the images of the chicken house acquired by several image collectors are configured as several images covering different ranges of the target chicken house.

[0024] Furthermore, the image acquisition and processing module specifically includes: an image segmentation unit; wherein the image segmentation unit is configured to use a PIDNet model trained based on chicken images to identify and segment chicken images in a plurality of chicken house images of a target chicken house, and obtain a chicken image detection result for each image; wherein the chicken image detection result is specifically: , , …, ;in, represents the chicken image detection result of the Mth chicken house image, The mask representing the Nth chicken detected in the Mth chicken house image.

[0025] Furthermore, the image acquisition and processing module specifically includes: a depth estimation unit; wherein the depth estimation unit is configured to use the SideRT model to extract the depth information of each chicken house image and store it in the form of a depth map.

[0026] In this embodiment, the chicken house images are collected by image sensors arranged at different positions of the target chicken house and covering different ranges of the target chicken house. After collecting several chicken house images, PIDNet (i.e., real-time semantic segmentation model) is used to segment several chicken house images (such as Figure 2As shown, in this embodiment, four images of the chicken house are collected by four image sensors A, B, C, and D arranged at the midpoints of the four sides of the rectangular chicken house. , , , ) and perform chicken image segmentation based on the recognition detection results to obtain a series of detection results under each image , , , , represent The test results, represent The mask of the first chicken detected is then extracted using the SideRT model (i.e., a Transformer structure model based on single image depth estimation) to extract the depth information of each chicken house image, which can then be used to measure the size of the chickens in the target chicken house and provide data support for estimation. This has higher accuracy than the traditional weight detection that is only achieved through image information. At the same time, this embodiment uses a method of collecting images of the interior of the area from the edge of the area of ​​the target chicken house, so that the placement of the image sensor is no longer constrained, thereby improving scene adaptability (the existing method requires the image sensor to shoot vertically downward, so the viewing angle shooting area is limited).

[0027] In a preferred embodiment, the size measurement module specifically includes: a 2D feature extraction unit; wherein the 2D feature extraction unit is configured to perform scaling processing on the information extraction target generated by the mask based on each chicken image based on the depth information, and calculate the 2D features of each chicken image; wherein the 2D features include: a projected area S, a width W of the chicken image, and a perimeter L of the chicken image, and the specific expression is: ; ; ; in, is the area of ​​the mask, is the width of the mask, is the perimeter of the mask, D is the average depth value of the mask location, k is a constant for adjusting the degree of nonlinearity, and b are scaling factors used to adjust the effect of depth on area, and c is an offset used to prevent the area from being scaled too large when the depth value is too small.

[0028] Furthermore, the size measurement module specifically includes: a 3D feature extraction unit; wherein the 3D feature extraction unit is configured to estimate the 3D features of each chicken image according to the depth information of each chicken house image and the projection area S; wherein the 3D features include: volume V, the expression is specifically: wherein the 3D features include: volume V, the expression is specifically: ; Where V is the estimated volume, represents the average depth value of the mask perimeter of each chicken image, Represents the depth value of the i-th pixel at the corresponding position of the mask of each chicken image.

[0029] In this embodiment, the 2D feature extraction unit uses the filled mask as the information extraction target and scales it using the depth information, thereby calculating and obtaining the 2D features of each chicken image. The 2D features include: the projection area S, the width W of the chicken image, and the perimeter L of the chicken image. The projection area is the result of scaling the mask area in combination with the depth information, the width of the chicken is the result of scaling the maximum width of the mask according to the depth, and the perimeter of the chicken is the result of scaling the outer line length of the mask according to the depth. After obtaining the 2D features of the chicken image, the volume of the chicken is estimated by the depth information of the corresponding area of ​​the mask to obtain the 3D features of the chicken image, that is, the volume parameters of the chicken. Compared with the traditional weight detection only realized by image information, the accuracy of the body shape and weight assessment of a large number of chickens in the chicken house is improved.

[0030] In a preferred embodiment, the weight estimation module specifically includes: a 3D model generation unit, which is configured to generate a feature data group based on the 2D features and 3D features of each chicken in each chicken house image, input the feature data group into the neural radiation field model to obtain a feature potential representation, and input the feature potential representation into the CLIP-Mesh model to obtain a 3D model of each chicken; and a volume correction unit, which is configured to correct the 3D features of each chicken to the volume corresponding to the 3D model according to the 3D model of each chicken.

[0031] Furthermore, the weight estimation module specifically includes: a weight calculation unit; wherein the weight calculation unit is configured to calculate the weight parameters of each chicken based on the 3D model of each chicken and a preset ratio of weight to unit volume.

[0032] In this embodiment, considering the use of a combination of image recognition and depth information to measure and evaluate the size and weight of chickens in the target chicken house, although it has certain advantages over traditional technologies, it still does not meet very high standards in terms of measurement accuracy. Therefore, after obtaining the 2D and 3D features of the chicken image based on image recognition and depth information technology, NeRF is used to generate a potential representation of the chicken-related information, and then CLIP-Mesh is used to obtain the corrected volume and calculate the weight according to preset rules. Specifically, the projection area S, the width W of the chicken, the circumference L of the chicken, and the volume V are first obtained through the 3D model generation unit to form data μ, and then the data μ is sent to the neural radiation field model (such as the MipNeRF model) to obtain the potential representation: , (θ is the angle between the image sensor and the horizontal plane, x is the potential representation), after which the potential representation x and the text "Chicken" are sent to CLIP-Mesh (a model structure based on CLIP and Mesh neural network) to obtain the 3D model of the chicken corresponding to the potential representation x. Finally, the volume parameters of the 3D model are used to replace the volume parameters previously calculated by 2D features and 3D features, and the weight parameters of each chicken are calculated by the 3D model in a "weight / unit volume" manner, thereby providing a highly accurate chicken body shape and weight measurement method.

[0033] In a preferred embodiment, the result feedback module specifically includes: an abnormality judgment unit, which is configured to judge whether the size parameter and / or weight parameter of each chicken is an abnormal value, and if so, judge the chicken as an abnormal target; an evaluation alarm unit, which is configured to judge whether the number of chickens judged as abnormal targets in each chicken house image collected at each moment exceeds a preset number, and if so, feedback an evaluation alarm signal.

[0034] In this embodiment, considering that in actual applications, there are a large number of chickens in each area of ​​the target chicken coop, if the target chicken coop is fully monitored and each chicken is detected and tracked, it will cause a lot of computing overhead and high equipment costs. Therefore, this embodiment abandons the idea of ​​full-area full-target tracking detection, and adopts a probability calculation based on the number of abnormal targets in a single area. It monitors the chicken house images collected by each image sensor at each moment to see whether the number of chickens with abnormal body shape and / or weight (body shape is measured by 2D features or volume parameters corrected by 3D models, and weight is measured by weight parameters calculated by 3D models) exceeds a preset number. If so, it indicates that there is a high possibility that there are chickens with abnormal body shape and weight in the target chicken coop, and the breeder needs to adjust and optimize the feeding strategy. Specifically, if Figure 3As shown, when there are unqualified chickens exceeding the threshold in the area captured by image sensor A during use, an alarm message is sent to the user, and the approximate range of the unqualified chickens is given according to the depth information. For example, if the depth value is 3m, the unqualified chickens are 3m away from the image sensor A. Therefore, the equipment computing load and cost budget are reduced while ensuring the accuracy of chicken body shape and weight monitoring and evaluation.

[0035] Embodiment 2: Reference Figure 4 , Figure 4 A schematic flow chart of a chicken body shape and weight assessment method based on image recognition provided in an embodiment of the present invention.

[0036] like Figure 4 As shown, a chicken body shape and weight assessment method based on image recognition is used in the chicken body shape and weight assessment system based on image recognition as described above, comprising: S1: Acquire several chicken house images of the target chicken house, identify and segment several chicken images in each chicken house image, and extract depth information in the chicken house image; S2: Based on the plurality of chicken images obtained by segmentation and the depth information, the size parameters of each chicken in each chicken house image are calculated; wherein the size parameters include 2D features and 3D features of each chicken; S3: Based on the 2D features and 3D features of each chicken in each chicken house image, a 3D model of each chicken is generated, the size parameters of each chicken are corrected, and the weight parameters of each chicken are calculated; S4: Determine whether there are chicken images with abnormal size parameters and / or weight parameters exceeding a preset number in each chicken house image collected at each moment, and if so, feedback evaluation alarm information.

[0037] This embodiment provides a chicken body shape and weight assessment method based on image recognition, by extracting several chicken images in the chicken house image, using the depth information of the chicken house image to measure the 2D features and 3D features of the chicken corresponding to each chicken image, and then using the 2D features and 3D features to generate a 3D model of each chicken and calculate the weight parameter, and finally according to the number of chicken images with abnormal size parameters and / or weight parameters in the several chicken house images, it is determined whether to feedback the assessment alarm signal. Thus, by combining image recognition with depth information, the position of the image sensor in the chicken house is no longer constrained, and the scene adaptability is improved. At the same time, by generating a 3D model of each chicken to correct the volume, and determining the weight parameter based on the conversion relationship between the unit volume and weight of the model, the accuracy of the body shape and weight assessment of a large number of chickens in the chicken house is improved.

[0038] The specific implementation of the chicken body shape and weight assessment method based on image recognition in the present application is basically the same as the above-mentioned chicken body shape and weight assessment system embodiment based on image recognition, and will not be repeated here.

[0039] In the description of the embodiments of the present invention, it should be understood that the terms "upper", "lower", "front", "back", "left", "right", "vertical", "horizontal", "center", "top", "bottom", "top", "bottom", "inside", "outside", "inner side", "outer side" and the like indicate orientations or positional relationships based on the orientations or positional relationships shown in the accompanying drawings, and are only for the convenience of describing the present invention and simplifying the description, rather than indicating or implying that the device or element referred to must have a specific orientation, be constructed and operated in a specific orientation, and therefore cannot be understood as a limitation on the present invention. Among them, "inside" refers to an internal or enclosed area or space. "Periphery" refers to the area surrounding a specific component or a specific area.

[0040] In the description of the embodiments of the present invention, the terms "first", "second", "third", and "fourth" are used for descriptive purposes only and are not to be understood as indicating or implying relative importance or implicitly indicating the number of the indicated technical features. Thus, a feature defined as "first", "second", "third", and "fourth" may explicitly or implicitly include one or more of the features. In the description of the present invention, unless otherwise specified, "multiple" means two or more.

[0041] In the description of the embodiments of the present invention, it should be noted that, unless otherwise clearly specified and limited, the terms "install", "connect", "connect", and "assemble" should be understood in a broad sense, for example, it can be a fixed connection, a detachable connection, or an integral connection; it can be a direct connection, or an indirect connection through an intermediate medium, or it can be the internal communication of two components. For ordinary technicians in this field, the specific meanings of the above terms in the present invention can be understood according to specific circumstances.

[0042] In the description of the embodiments of the present invention, specific features, structures, materials or characteristics may be combined in a suitable manner in any one or more embodiments or examples.

[0043] In the description of the embodiments of the present invention, it should be understood that "-" and "~" represent a range between two values, and the range includes the endpoints. For example: "AB" represents a range greater than or equal to A and less than or equal to B. "A~B" represents a range greater than or equal to A and less than or equal to B.

[0044] In the description of the embodiments of the present invention, the term "and / or" herein is merely a description of the association relationship of associated objects, indicating that three relationships may exist. For example, A and / or B may represent: A exists alone, A and B exist at the same time, and B exists alone. In addition, the character " / " herein generally indicates that the associated objects before and after are in an "or" relationship.

[0045] Although embodiments of the present invention have been shown and described, it will be appreciated by those skilled in the art that various changes, modifications, substitutions and variations may be made to the embodiments without departing from the principles and spirit of the present invention, and that the scope of the present invention is defined by the appended claims and their equivalents.

Claims

1. A chicken body shape and weight assessment system based on image recognition, characterized in that: The system comprises: An image acquisition and processing module, wherein the image acquisition and processing module is configured to acquire a plurality of chicken house images of a target chicken house, identify and segment a plurality of chicken images in each chicken house image, and extract depth information from the chicken house images; A size calculation module, wherein the size calculation module is configured to calculate the size parameters of each chicken in each chicken house image based on the plurality of chicken images obtained by segmentation and the depth information; wherein the size parameters include 2D features and 3D features of each chicken; A weight estimation module, wherein the weight estimation module is configured to generate a 3D model of each chicken based on the 2D features and 3D features of each chicken in each chicken house image, correct the size parameters of each chicken and calculate the weight parameters of each chicken; The result feedback module is configured to determine whether there are chicken images with abnormal size parameters and / or weight parameters exceeding a preset number in each chicken house image collected at each moment, and if so, feedback evaluation alarm information.

2. The chicken body shape and weight assessment system based on image recognition according to claim 1, characterized in that: The image acquisition and processing module specifically includes: Image acquisition unit; Wherein, the image acquisition unit is configured to acquire images of the chicken house acquired by a plurality of image sensors arranged at different positions of the target chicken house; The chicken house images collected by several image collectors are configured as several images covering different ranges of the target chicken house.

3. The chicken body shape and weight assessment system based on image recognition according to claim 1, characterized in that: The image acquisition and processing module specifically includes: Image segmentation unit; The image segmentation unit is configured to use a PIDNet model trained based on chicken images to identify and segment chicken images in a plurality of chicken house images of a target chicken house, and obtain a chicken image detection result for each image; The chicken image detection result is specifically: , , …, ; in, represents the chicken image detection result of the Mth chicken house image, The mask representing the Nth chicken detected in the Mth chicken house image.

4. The chicken body shape and weight assessment system based on image recognition according to claim 3, characterized in that: The image acquisition and processing module specifically includes: Depth estimation unit; The depth estimation unit is configured to extract the depth information of each chicken house image using the SideRT model and store it in the form of a depth map.

5. The chicken body shape and weight assessment system based on image recognition according to claim 3, characterized in that: The size measurement module specifically includes: 2D feature extraction unit; The 2D feature extraction unit is configured to perform scaling processing on the information extraction target generated by the mask based on each chicken image based on the depth information, and calculate and obtain the 2D features of each chicken image; The 2D features include: the projected area S, the width W of the chicken image, and the perimeter L of the chicken image, and the specific expression is: ； ； ； in, is the area of ​​the mask, is the width of the mask, is the perimeter of the mask, D is the average depth value of the mask location, k is a constant for adjusting the degree of nonlinearity, and b are scaling factors used to adjust the effect of depth on area, and c is an offset used to prevent the area from being scaled too large when the depth value is too small.

6. The chicken body shape and weight assessment system based on image recognition according to claim 5, characterized in that: The size measurement module specifically includes: 3D feature extraction unit; Wherein, the 3D feature extraction unit is configured to estimate the 3D features of each chicken image according to the depth information of each chicken house image and the projection area S; The 3D features include: volume V, which is specifically expressed as: ； Where V is the estimated volume, represents the average depth value of the mask perimeter of each chicken image, Represents the depth value of the i-th pixel at the corresponding position of the mask of each chicken image.

7. The chicken body shape and weight assessment system based on image recognition according to claim 6, characterized in that: The weight estimation module specifically includes: A 3D model generation unit, wherein the 3D model generation unit is configured to generate a feature data set based on the 2D features and 3D features of each chicken in each chicken house image, input the feature data set into a neural radiation field model to obtain a feature potential representation, and input the feature potential representation into a CLIP-Mesh model to obtain a 3D model of each chicken; A volume correction unit is configured to correct the 3D features of each chicken into a volume corresponding to the 3D model according to the 3D model of each chicken.

8. The chicken body shape and weight assessment system based on image recognition according to claim 7, characterized in that: The weight estimation module specifically includes: Weight calculation unit; The weight calculation unit is configured to calculate the weight parameters of each chicken based on the 3D model of each chicken and a preset ratio of weight to unit volume.

9. The chicken body shape and weight assessment system based on image recognition according to claim 8, characterized in that: The result feedback module specifically includes: An abnormality determination unit, the abnormality determination unit is configured to determine whether the size parameter and / or weight parameter of each chicken is an abnormal value, and if so, determine the chicken as an abnormal target; An evaluation alarm unit is configured to determine whether the number of chickens determined to be abnormal targets in each chicken house image collected at each moment exceeds a preset number, and if so, feed back an evaluation alarm signal.

10. A chicken body shape and weight assessment method based on image recognition, characterized in that: The method comprises: S1: Acquire several chicken house images of the target chicken house, identify and segment several chicken images in each chicken house image, and extract depth information in the chicken house image; S2: Based on the plurality of chicken images obtained by segmentation and the depth information, the size parameters of each chicken in each chicken house image are calculated; wherein the size parameters include 2D features and 3D features of each chicken; S3: Based on the 2D features and 3D features of each chicken in each chicken house image, a 3D model of each chicken is generated, the size parameters of each chicken are corrected, and the weight parameters of each chicken are calculated; S4: Determine whether there are chicken images with abnormal size parameters and / or weight parameters exceeding a preset number in each chicken house image collected at each moment, and if so, feedback evaluation alarm information.

Images