Collapse measurement method and system based on image analysis

Abstract

The invention discloses a cockscomb measurement method and system based on image analysis. The method comprises the steps that a large number of cockscomb images are collected to serve as original images; preprocessing the original image to obtain a processed image; performing feature extraction on the processed image to obtain a feature vector; and based on the feature vectors, constructing an image analysis model to complete the measurement of the cockscomb. According to the method, accurate measurement of the cockscomb lacking obvious characteristics and having large individual difference is realized, rich geometric information is provided for subsequent morphological characteristic analysis of the cockscomb, and the method has important application value in the fields of livestock breeding, animal genetic breeding and the like.

Description

Technical Field

[0001] The present invention relates to the field of image processing, and in particular to a chicken comb measurement method and system based on image analysis. Background Art

[0002] In the process of comb measurement, traditional manual measurement methods have problems such as low efficiency, low accuracy, and incomplete measurement data. In order to obtain more abundant and accurate comb morphological feature data, it is necessary to introduce a measurement method based on image analysis. However, in practical applications, there are a large number of wrinkles and uneven areas on the comb surface, which leads to a large number of interference factors such as occlusion, shadow and reflection in the collected image data, affecting the accuracy of subsequent image analysis and three-dimensional reconstruction. At the same time, there are no obvious feature points on the comb surface, and there is a lack of reliable matching relationship between images collected at different angles, which brings challenges to the registration and fusion of multi-view images. In addition, there are large individual differences in comb morphology. The combs of different breeds and different growth stages have significant differences in size, color and shape, which makes it difficult to establish a unified analysis model and evaluation criteria based on image analysis, affecting the comparability and interpretability of the measurement results. Summary of the invention

[0003] In order to solve the technical problems in the above background, the present invention aims to propose a feature-based image recognition method and system for solving the problem of lack of obvious feature points on the surface of chicken comb.

[0004] To achieve the above object, the present invention provides a method for measuring chicken combs based on image analysis, the steps comprising:

[0005] Collect a large number of cockscomb images as original images;

[0006] Preprocessing the original image to obtain a processed image;

[0007] Performing feature extraction on the processed image to obtain a feature vector;

[0008] Based on the feature vector, an image analysis model is constructed to complete the measurement of the comb.

[0009] Preferably, the method for performing the pretreatment comprises:

[0010] Using a wavelet denoising method to remove Gaussian noise and salt and pepper noise in the original image;

[0011] Image restoration is performed on the original image from which Gaussian noise and salt and pepper noise are removed to obtain the processed image.

[0012] Preferably, the method for performing the feature extraction includes:

[0013] Based on the processed image, generating a binary image;

[0014] Based on the binary image, a region including the cockscomb is obtained;

[0015] Based on the cockscomb region, the feature vector is extracted.

[0016] Preferably, the method for completing the measurement of the comb includes: inputting the extracted feature vector into a pre-trained support vector machine model, classifying the feature vector through the support vector machine model, identifying the comb area, and completing the measurement of the comb.

[0017] The present invention also provides a chicken comb measurement system based on image analysis, the system is used to implement the above method, including: an acquisition module, a processing module, an extraction module and a measurement module;

[0018] The acquisition module is used to acquire a large number of cockscomb images as original images;

[0019] The processing module is used to pre-process the original image to obtain a processed image;

[0020] The extraction module is used to extract features from the processed image to obtain a feature vector;

[0021] The measurement module is used to construct an image analysis model based on the feature vector to complete the measurement of the chicken comb.

[0022] Preferably, the workflow of the processing module includes:

[0023] Using a wavelet denoising method to remove Gaussian noise and salt and pepper noise in the original image;

[0024] Image restoration is performed on the original image from which Gaussian noise and salt and pepper noise are removed to obtain the processed image.

[0025] Preferably, the workflow of the extraction module includes:

[0026] Based on the processed image, generating a binary image;

[0027] Based on the binary image, a region including the cockscomb is obtained;

[0028] Based on the cockscomb region, the feature vector is extracted.

[0029] Preferably, the workflow of the measurement module includes: inputting the extracted feature vector into a pre-trained support vector machine model, classifying the feature vector through the support vector machine model, identifying the comb area, and completing the measurement of the comb.

[0030] Compared with the prior art, the present invention has the following beneficial effects:

[0031] The present invention realizes the accurate measurement of combs that lack obvious features and have large individual differences, provides rich geometric information for subsequent comb morphological feature analysis, and has important application value and broad application prospects in the fields of animal husbandry, animal genetic breeding, etc. BRIEF DESCRIPTION OF THE DRAWINGS

[0032] In order to more clearly illustrate the technical solution of the present invention, the following briefly introduces the drawings required for use in the embodiments. 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 paying creative labor.

[0033] Figure 1 The figure is a schematic diagram of a method flow of an embodiment of the present invention. DETAILED DESCRIPTION

[0034] 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.

[0035] In order to make the above-mentioned objects, features and advantages of the present invention more obvious and easy to understand, the present invention is further described in detail below with reference to the accompanying drawings and specific embodiments.

[0036] like Figure 1 FIG. 1 is a schematic diagram of the method flow of this embodiment, and the steps include:

[0037] S1. Collect a large number of chicken comb images as original images.

[0038] A large number of clear comb images are collected as original images from open source poultry websites, official websites of large farms, or national agricultural websites. These images can be images of normally developed combs, or images of underdeveloped or diseased combs.

[0039] S2. Preprocess the original image to obtain a processed image.

[0040] According to the acquired original image, the image preprocessing technology is used to denoise, enhance and correct the image, eliminate the interference introduced by factors such as lighting and occlusion, and obtain higher quality cockscomb image data, that is, the processed image.

[0041] Specifically, the wavelet denoising method is used to remove Gaussian noise and salt and pepper noise in the original image. First, a suitable wavelet basis function (db4 in this embodiment) and the number of decomposition layers (3 layers in this embodiment) are selected to perform wavelet decomposition on the image, and the high-frequency coefficients are soft-thresholded in the wavelet domain to remove noise, and then wavelet reconstruction is performed to obtain the denoised cockscomb image. The denoised cockscomb image is repaired, and a repair algorithm based on local texture synthesis is used for the missing areas and broken edges in the image. By selecting texture blocks similar to the missing areas in the image, the missing pixels are filled in a weighted average manner to maintain the consistency of the repaired area with the surrounding texture. The cockscomb images after geometric correction, brightness correction, image enhancement, image denoising, and image restoration are fused, and the principal component analysis method is used to fuse multiple processing results. By performing feature decomposition on the image, extracting the principal components, and then weighted combining the principal components, the final high-quality cockscomb image, that is, the processed image, is obtained.

[0042] S3. Perform feature extraction on the processed image to obtain a feature vector.

[0043] A method based on color threshold segmentation is adopted. First, the processed image is converted into HSV color space, and then the threshold range of the three channels H, S, and V is set according to the color characteristics of the chicken's comb, and the pixels that meet the threshold conditions are extracted to generate a binary image. By performing a connected domain analysis on the binary image, the connected domain with the largest area is retained as the chicken's comb area, and an image containing the chicken's comb is obtained. For the extracted chicken's comb area image, the Canny operator is used for edge detection, and the contour features of the chicken's comb are extracted by setting high and low thresholds (in this embodiment, set to 3 and 7). According to the contour features, the chicken's comb area is morphologically opened, and a 3×3 rectangular structural element is selected, and it is iterated twice to remove the noise caused by uneven illumination. Then the denoised image is closed, and a 5×5 rectangular structural element is selected, and it is iterated once to fill the contour breaks caused by occlusion. By calculating the change rate of the number of pixels of the chicken's comb contour before and after processing, if the change rate is less than 5%, it is considered that the interference has been effectively eliminated. Otherwise, the size of the structural element and the number of iterations are adaptively adjusted according to the change rate until the conditions are met. Finally, the enhanced contour features are extracted and the seven Hu invariant moments of the contour are calculated to form a feature vector.

[0044] S4. Based on the feature vector, an image analysis model is constructed to complete the measurement of the comb.

[0045] The feature vector of the extracted high-quality comb image is input into the pre-trained support vector machine model, and the feature vector is classified by the support vector machine model to identify the comb area, and the identification result is used for subsequent poultry health assessment. By analyzing the color, size, shape and other characteristics of the comb, combined with the poultry growth cycle, the health level and growth and development of the poultry can be assessed.

[0046] Embodiment 2

[0047] Firstly, a large number of cockscomb images are collected as original images using the acquisition module.

[0048] A large number of clear comb images are collected as original images from open source poultry websites, official websites of large farms, or national agricultural websites. These images can be images of normally developed combs, or images of underdeveloped or diseased combs.

[0049] The processing module preprocesses the original image to obtain a processed image.

[0050] According to the acquired original image, the image preprocessing technology is used to denoise, enhance and correct the image, eliminate the interference introduced by factors such as lighting and occlusion, and obtain higher quality cockscomb image data, that is, the processed image.

[0051] Specifically, the wavelet denoising method is used to remove Gaussian noise and salt and pepper noise in the original image. First, a suitable wavelet basis function (db4 in this embodiment) and the number of decomposition layers (3 layers in this embodiment) are selected to perform wavelet decomposition on the image, and the high-frequency coefficients are soft-thresholded in the wavelet domain to remove noise, and then wavelet reconstruction is performed to obtain the denoised cockscomb image. The denoised cockscomb image is repaired, and a repair algorithm based on local texture synthesis is used for the missing areas and broken edges in the image. By selecting texture blocks similar to the missing areas in the image, the missing pixels are filled in a weighted average manner to maintain the consistency of the repaired area with the surrounding texture. The cockscomb images after geometric correction, brightness correction, image enhancement, image denoising, and image restoration are fused, and the principal component analysis method is used to fuse multiple processing results. By performing feature decomposition on the image, extracting the principal components, and then weighted combining the principal components, the final high-quality cockscomb image, that is, the processed image, is obtained.

[0052] The extraction module extracts features from the processed image to obtain a feature vector.

[0053] A method based on color threshold segmentation is adopted. First, the processed image is converted into HSV color space, and then the threshold range of the three channels H, S, and V is set according to the color characteristics of the chicken's comb, and the pixels that meet the threshold conditions are extracted to generate a binary image. By performing a connected domain analysis on the binary image, the connected domain with the largest area is retained as the chicken's comb area, and an image containing the chicken's comb is obtained. For the extracted chicken's comb area image, the Canny operator is used for edge detection, and the contour features of the chicken's comb are extracted by setting high and low thresholds (in this embodiment, set to 3 and 7). According to the contour features, the chicken's comb area is morphologically opened, and a 3×3 rectangular structural element is selected, and it is iterated twice to remove the noise caused by uneven illumination. Then the denoised image is closed, and a 5×5 rectangular structural element is selected, and it is iterated once to fill the contour breaks caused by occlusion. By calculating the change rate of the number of pixels of the chicken's comb contour before and after processing, if the change rate is less than 5%, it is considered that the interference has been effectively eliminated. Otherwise, the size of the structural element and the number of iterations are adaptively adjusted according to the change rate until the conditions are met. Finally, the enhanced contour features are extracted and the seven Hu invariant moments of the contour are calculated to form a feature vector.

[0054] The measurement module constructs an image analysis model based on the feature vector to complete the measurement of the comb.

[0055] The feature vector of the extracted high-quality comb image is input into the pre-trained support vector machine model, and the feature vector is classified by the support vector machine model to identify the comb area, and the identification result is used for subsequent poultry health assessment. By analyzing the color, size, shape and other characteristics of the comb, combined with the poultry growth cycle, the health level and growth and development of the poultry can be assessed.

[0056] The embodiments described above are only descriptions of the preferred embodiments of the present invention and are not intended to limit the scope of the present invention. Without departing from the design spirit of the present invention, various modifications and improvements made to the technical solutions of the present invention by ordinary technicians in this field should all fall within the protection scope determined by the claims of the present invention.

Claims

1. A method for measuring chicken combs based on image analysis, characterized in that the steps include: Collect a large number of cockscomb images as original images; Preprocessing the original image to obtain a processed image; Performing feature extraction on the processed image to obtain a feature vector; Based on the feature vector, an image analysis model is constructed to complete the measurement of the comb.

2. The method for measuring chicken combs based on image analysis according to claim 1, characterized in that: The method for performing the pretreatment comprises: Using a wavelet denoising method to remove Gaussian noise and salt and pepper noise in the original image; Image restoration is performed on the original image from which Gaussian noise and salt and pepper noise are removed to obtain the processed image.

3. The method for measuring cockscombs based on image analysis according to claim 1, characterized in that: The method for performing the feature extraction comprises: Based on the processed image, generating a binary image; Based on the binary image, a region including the cockscomb is obtained; Based on the cockscomb region, the feature vector is extracted.

4. The method for measuring chicken combs based on image analysis according to claim 1, characterized in that: The method for completing the measurement of the comb includes: inputting the extracted feature vector into a pre-trained support vector machine model, classifying the feature vector through the support vector machine model, identifying the comb area, and completing the measurement of the comb.

5. A chicken comb measurement system based on image analysis, the system is used to implement the method according to any one of claims 1 to 4, characterized in that: include: Acquisition module, processing module, extraction module and measurement module; The acquisition module is used to acquire a large number of cockscomb images as original images; The processing module is used to pre-process the original image to obtain a processed image; The extraction module is used to extract features from the processed image to obtain a feature vector; The measurement module is used to construct an image analysis model based on the feature vector to complete the measurement of the chicken comb.

6. The comb measurement system based on image analysis according to claim 5, characterized in that: The workflow of the processing module includes: Using a wavelet denoising method to remove Gaussian noise and salt and pepper noise in the original image; Image restoration is performed on the original image from which Gaussian noise and salt and pepper noise are removed to obtain the processed image.

7. The comb measurement system based on image analysis according to claim 5, characterized in that: The workflow of the extraction module includes: Based on the processed image, generating a binary image; Based on the binary image, a region including the cockscomb is obtained; Based on the cockscomb region, the feature vector is extracted.

8. The comb measurement system based on image analysis according to claim 5, characterized in that: The working process of the measurement module includes: inputting the extracted feature vector into a pre-trained support vector machine model, classifying the feature vector through the support vector machine model, identifying the comb area, and completing the measurement of the comb.

Images