A training method for an intelligent detection model of palisade cells in Cuscuta australis
Through deep learning technology, the intelligent detection model of southern Cuscuta junction-shaped cells was trained, and the problem of lack of talents for identification of Chinese medicinal materials was solved, efficient and accurate identification and detection of Chinese medicinal materials was achieved, and the application of artificial intelligence technology in detection and auxiliary identification of Chinese medicinal materials was promoted.
Patent Information
- Application Number
- CN202411815016.8
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Filing Date
- 2024-12-11
- Publication Date
- 2025-08-08
- Estimated Expiration
- 2044-12-11
AI Technical Summary
There is a shortage of talents in identifying authenticity of Chinese medicinal materials, and the traditional identification efficiency is low, which cannot meet the needs of fast and efficient Chinese medicine inspection.
Deep learning technology is used to train the intelligent detection model of southern dodder tray cells. Through feature annotation, irregular aggregation, cross-training and iterative optimization, an efficient intelligent detection model of southern dodder tray cells is constructed.
It improves the accuracy and efficiency of identification and testing of traditional Chinese medicinal materials, reduces the need for manual intervention, and adapts to the needs of modern biological research and clinical applications.
Smart Images
Figure CN119863791B_ABST
Abstract
Description
Technical Field
[0001] The present application relates to the technical field of intelligent processing and analysis of microscopic imaging of traditional Chinese medicine, and in particular to a method for training an intelligent detection model of palisade cells of Cuscuta australis. Background Art
[0002] With the sweeping of the global information wave, intelligence has become a major trend in the development of various industries. At present, the supervision of traditional Chinese medicine and ethnic medicine is constantly increasing, the inspection tasks are increasing year by year, and there is an extreme shortage of traditional identification talents. It is urgent to carry out research on intelligent identification methods of traditional Chinese medicine and ethnic medicine based on digital images of traditional Chinese medicine specimens and the application of cutting-edge technologies such as computer image recognition to fill the gaps in the industry and lay the foundation for the development of new scientific supervision tools.
[0003] Authenticity identification of Chinese medicinal materials is an important guarantee for the safety and effectiveness of Chinese medicine; however, with the continuous increase in the supervision of Chinese medicine, the workload of Chinese medicine inspection is also gradually increasing. Therefore, a fast and efficient work pace has become the daily routine of Chinese medicine inspectors; currently there is an extreme shortage of talents in the authenticity identification of Chinese medicinal materials. The mastery of the authenticity identification of Chinese medicinal materials needs to be based on long-term practical accumulation and accurate observation of a large number of accurate samples. It is urgent to carry out key technology research on the intelligent application of Chinese medicine ethnic specimens to serve daily inspection and grassroots supervision.
[0004] Compared with traditional manual identification, the use of artificial intelligence detection models can greatly reduce the manpower required for identification work, improve the work efficiency of identification personnel, and solve the problems of lack of identification talents and high work pressure of traditional Chinese medicine inspectors; this method uses digital microscopic imaging of the palisade cells of southern dodder and applies image recognition technologies such as convolutional networks. By training feature extraction and feature recognition modules, a deep convolutional neural network model with high reliability and good generalization performance is obtained, and the palisade cells of southern dodder are identified in the application.
[0005] In view of this, the present application is proposed to provide a method for training an intelligent detection model for palisade cells of Cuscuta australis that overcomes the above-mentioned problem or at least partially solves the above-mentioned problem. Summary of the Invention
[0006] In order to solve the above problems, the present invention discloses a method for training an intelligent detection model of palisade cells of Cuscuta australis, comprising:
[0007] S1. Collect digital microscopic images of palisade cells of Cuscuta australis, each of which has feature annotations; randomly aggregate the digital microscopic images based on the feature annotations to obtain A microscopic image clusters, and split the microscopic image clusters into clusters to obtain training clusters and optimization clusters;
[0008] S2. Build an initial intelligent detection model for palisade cells of Cuscuta australis based on deep learning technology;
[0009] S3, cross-training the initial southern dodder palisade cell intelligent detection model using the training cluster to obtain the optimal southern dodder palisade cell intelligent detection model;
[0010] S4. Use the optimization cluster to iteratively optimize the most suitable southern dodder palisade cell intelligent detection model to obtain the target southern dodder palisade cell intelligent detection model.
[0011] Furthermore, the feature annotation includes: feature classification annotation and target positioning annotation; target positioning annotation includes: marking position coordinates on digital microscopic imaging; feature classification annotation includes: marking image features of palisade cells of southern dodder on digital microscopic imaging.
[0012] Furthermore, the method of randomly aggregating digital microscopic imaging includes:
[0013] Construct microscopic feature vectors based on feature annotation, use similarity calculation formula to calculate the similarity between different microscopic feature vectors, and construct similarity matrix based on similarity ;by Represents the similarity matrix Rank The elements of the column, Representative The microscopic eigenvectors and The similarity between the microscopic eigenvectors; for the similarity matrix The index of the column whose element value is greater than the preset similarity threshold constitutes the first The adjacent set of microscopic feature vectors is constructed; the fitness of each microscopic feature vector is calculated based on the similarity matrix and the adjacent set. The fitness calculation formula is: ;in, Representative The local fitness of the microscopic eigenvectors, Represents an approaching collection, represents the size of the adjacent set, represents the adjustment parameter; cluster analysis is performed through local fitness, similar digital microscopy images are classified into one category, and A microscopy clusters are obtained, where A is an integer greater than zero.
[0014] Furthermore, the construction method of the initial intelligent detection model of palisade cells of Cuscuta australis includes:
[0015] Initialize the feature extraction model as a neural network model. The network layer structure includes: input layer, convolution layer, pooling layer, fully connected layer and output layer. Sigmoid function is used as activation function. The input layer uses the training set as input, and the output layer uses the feature cell map as output.
[0016] The feature recognition model is initialized as the Faster R-CNN model, and the region proposal network is defined to use the Anchor mechanism as the candidate box generation mechanism. The classifier is the SoftMax classifier, the regressor is the Lasso regressor, the feature cell map is used as the input of the feature recognition model, and the target recognition result is used as the output of the feature recognition model.
[0017] Furthermore, the method of cross-training the initial Cuscuta australis palisade cell intelligent detection model includes:
[0018] Based on the total number K of clusters in the training cluster, K model trainings are performed. Each time the model training is performed, one of the training clusters is used as the validation set, and the other training clusters are used as the training set for training, to obtain K initial southern dodder palisade cell intelligent detection models; the initial southern dodder palisade cell intelligent detection model includes a feature extraction model and a feature recognition model. For each model training, the digital microscopic imaging in the training set is used as the input of the feature extraction model, and the digital microscopic imaging in the training set is convolved through the convolution layer to obtain a feature map; the feature map is pooled through the pooling layer to obtain a low-dimensional feature map; the low-dimensional feature map is feature mapped through the fully connected layer to obtain a feature cell map; the feature recognition model is used to perform feature recognition on the feature cell map to obtain a target recognition result, and the target recognition result is confidence calculated to obtain a confidence parameter; the average value of the confidence parameters obtained from each training is calculated, and the initial southern dodder palisade cell intelligent detection model with the largest average value is used as the optimal southern dodder palisade cell intelligent detection model.
[0019] Furthermore, the method of performing a convolution operation on the digital microscopic images in the training set includes: presetting a convolution kernel movement stride and a convolution matrix size, sliding the convolution kernel of the convolution layer in the digital microscopic images at the movement stride, and multiplying and summing pixel elements of a selected digital microscopic imaging area having the same size as the convolution matrix by the convolution kernel after each slide, thereby extracting local features of the digital microscopic images;
[0020] The method of performing the pooling operation on the feature map includes: dynamically adjusting the pooling window based on the size of the feature map and the size of the digital microscopic image, and the length of the pooling window is calculated as follows:
[0021] ;in, represents the length of the pooling window, represents the length of digital microscopy imaging, represents the length of the feature map, Represents the floor function; the width of the pooling window is calculated as: ;in, Represents the width of the pooling window, represents the width of digital microscopy imaging, Represents the width of the feature map; select elements in the feature map according to the length and width of the pooling window. When the length of the pooling window is equal to the width, take the average value of the elements in the pooling window. When the length of the pooling window is not equal to the width, take the maximum value of the elements in the pooling window.
[0022] Furthermore, the method of performing feature recognition on the characteristic cell map includes:
[0023] The region proposal network scans the feature cell map and generates candidate boxes of different scales and aspect ratios using the anchor mechanism. The position and size of the bounding box are adjusted through the regressor. The classifier is used to identify the object and category of each candidate box. The candidate boxes identified as having the same category are then spliced together to obtain a new candidate box. Based on the size of the feature cell map and the size of the new candidate box, the area selected by the candidate box is converted into regional data. The converted regional data is the target recognition result.
[0024] Furthermore, the formula for calculating the confidence of the target recognition result is:
[0025] ;in Representative The confidence parameter of the target recognition result, Represents the total number of categories of identified content in the target recognition results, Represents the number of categories in the target recognition result that are consistent with the feature annotation content, Representative The target recognition results are the same as those of the The first feature annotation The correlation between categories, Represents the target recognition result The weight of each category, Representative The target recognition result categories, Representative The first feature annotation categories.
[0026] Furthermore, the method of iteratively optimizing the optimal palisade cell intelligent detection model for Cuscuta australis includes:
[0027] A genetic algorithm is used to perform convolution optimization on the convolution layer in the feature extraction model to obtain the optimal convolution group. The optimal convolution group is used as the convolution kernel of the convolution layer in the feature extraction model to obtain the target feature extraction model. The digital microscopic imaging in the optimized cluster is used as the input of the target feature extraction model for feature extraction to obtain the characteristic cell maps corresponding to different optimized clusters. The cell map features corresponding to the same optimized cluster constitute a feature map cluster. The feature map cluster is used as the input of the feature recognition model for feature recognition to obtain the target recognition result. The target recognition result is trained by a comprehensive loss function to supervise the entire learning process. The non-maximum suppression algorithm is used to suppress the repeated results of different candidate boxes to obtain the target feature recognition model.
[0028] Furthermore, the method of performing convolution optimization on the convolution layer in the feature extraction model includes:
[0029] For the convolutional layer in the feature extraction model, the default Initial convolution kernel groups, each convolution kernel group has E convolution kernels, initialize the optimal convolution group to be empty, initialize the weight parameters in each convolution kernel, and randomly take values based on the normal distribution L(a, b); based on the convolution kernel in the convolution layer, use the digital microscopy imaging in the optimized cluster as the input of the feature extraction model for forward propagation, use the output of the previous convolution layer as the input of the next convolution layer, use the cross entropy function as the loss function of each convolution layer, traverse the convolution kernel group, use each convolution kernel group as the convolution kernel in the convolution layer, and use the inverse of the loss function value as the convolution fitness of each convolution kernel group; calculate the selection probability of each convolution kernel group based on the convolution fitness, and the calculation formula of the selection probability is: ;in, Representative The selection probability of a convolution kernel group, Represents the index of the convolution kernel group, Representative The convolution fitness of the convolution kernel group, Representative The convolution fitness of the convolution kernel group; iterative selection is performed based on the selection probability, and F convolution kernel groups are selected as the parent convolution kernel groups each time. The parent convolution kernel groups are combined in pairs using the permutation and combination method to obtain the parent convolution pairs. For each pair of parent convolution pairs, the convolution kernel with the largest modulus value in the parent convolution kernel group is selected as the crossover node for single-point crossover to obtain the child convolution kernel group; the increment value is preset, the convolution kernels in the child convolution kernel group are traversed, and the weight parameters in the convolution kernel are selected and mutated using the random selection algorithm. Each weight parameter selected and mutated is consistent with the increment value. The values are added to obtain the mutated convolution kernel group; the mutated convolution kernel group is used as the convolution kernel in the convolution layer, and the convolution fitness of the mutated convolution kernel group is calculated. When the optimal convolution group is empty, the mutated convolution kernel group with the largest convolution fitness value is selected as the new optimal convolution group. When the optimal convolution group is not empty, the mutated convolution kernel group with the largest convolution fitness value that is greater than the convolution fitness value of the optimal convolution group is selected as the new optimal convolution group. Repeat until the convolution fitness of the optimal convolution group no longer changes, and output the optimal convolution group at this time as the convolution kernel of the convolution layer in the feature extraction model.
[0030] The technical effects and advantages of the present invention's method for training an intelligent detection model for palisade cells of Cuscuta australis are as follows:
[0031] The present invention utilizes artificial intelligence technologies such as deep learning to assist Chinese herbal medicine quality inspectors in conducting intelligent detection of palisade cells of southern dodder, thereby improving the accuracy and efficiency of screening work, and has significant application prospects. By combining deep learning technology with knowledge of Chinese medicinal material identification and detection, an intelligent detection technology for palisade cells of southern dodder based on digital microscopy is realized, which further improves the accuracy and efficiency of Chinese medicinal material identification and detection, effectively solves the defects in the detection process of palisade cells of southern dodder, and promotes the application and development of a new generation of artificial intelligence technology in intelligent detection and auxiliary identification of Chinese medicinal materials. Through an efficient automated training process, the need for manual intervention is reduced, the level of intelligent detection is improved, and the needs of modern biological research and clinical applications are met. BRIEF DESCRIPTION OF THE DRAWINGS
[0032] In order to more clearly illustrate the technical solution of the present application, the following is a brief introduction to the drawings required for the description of the present application. Obviously, the drawings described below are only some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without any creative work.
[0033] Figure 1 This is a flowchart of the steps of a method for training an intelligent detection model for palisade cells of Cuscuta australis provided in one embodiment of the present application;
[0034] Figure 2This is a schematic diagram of a network structure of a feature extraction model for an intelligent detection model of palisade cells of Cuscuta australis provided in one embodiment of the present application;
[0035] Figure 3 This is a schematic diagram of the network structure of a feature recognition model of an intelligent detection model for palisade cells of Cuscuta australis provided in one embodiment of the present application. DETAILED DESCRIPTION
[0036] To make the objectives, features, and advantages of this application more readily apparent, the present application is further described below in conjunction with the accompanying drawings and specific embodiments. It is apparent that the embodiments described are only a portion of the embodiments of this application, not all of them. All other embodiments derived by persons of ordinary skill in the art based on the embodiments in this application without inventive effort are also within the scope of protection of this application.
[0037] It should be noted that in any embodiment of the present invention, a method for training a model for intelligent detection of palisade cells of southern dodder is applied to intelligent detection of palisade cells of southern dodder through digital microscopic imaging of palisade cells of southern dodder. Confidence parameters, including the confidence interval of a probability sample, are interval estimates of a certain population parameter of this sample. The confidence interval shows the degree to which the true value of this parameter has a certain probability of falling around the measurement result. The confidence interval gives the range of credibility of the measured value of the measured parameter, that is, the "certain probability". This probability is called a confidence parameter.
[0038] Reference Figure 1 , shows a method for training an intelligent detection model for palisade cells of Cuscuta australis provided in one embodiment of the present application, comprising:
[0039] S1. Collect digital microscopic images of palisade cells of Cuscuta australis, each of which has feature annotations; randomly aggregate the digital microscopic images based on the feature annotations to obtain A microscopic image clusters, and split the microscopic image clusters into clusters to obtain training clusters and optimization clusters;
[0040] S2. Build an initial intelligent detection model for palisade cells of Cuscuta australis based on deep learning technology;
[0041] S3, cross-training the initial southern dodder palisade cell intelligent detection model using the training cluster to obtain the optimal southern dodder palisade cell intelligent detection model;
[0042] S4. Use the optimization cluster to iteratively optimize the most suitable southern dodder palisade cell intelligent detection model to obtain the target southern dodder palisade cell intelligent detection model.
[0043] Feature annotation includes feature classification annotation and target positioning annotation of the microscopic observation morphologies of three types of southern dodder palisade cells; wherein, feature classification annotation includes annotating the image features of the microscopic observation morphologies of the three types of southern dodder palisade cells on the digital microscopic imaging; target positioning annotation includes target positioning annotation of the image features on the digital microscopic imaging; wherein, the microscopic observation morphologies of the three types of southern dodder palisade cells include: surface view of southern dodder palisade cells, complete side view of southern dodder palisade cells, and incomplete side view of southern dodder palisade cells.
[0044] Methods for random aggregation of digital microscopy images include:
[0045] Construct microscopic feature vectors based on feature annotation, use similarity calculation formula to calculate the similarity between different microscopic feature vectors, and construct similarity matrix based on similarity ;by Represents the similarity matrix Rank The elements of the column, Representative The microscopic eigenvectors and The similarity between the microscopic eigenvectors; for the similarity matrix The index of the column whose element value is greater than the preset similarity threshold constitutes the first The adjacent set of microscopic feature vectors is constructed; the fitness of each microscopic feature vector is calculated based on the similarity matrix and the adjacent set. The fitness calculation formula is: ;in, Representative The local fitness of the microscopic eigenvectors, Represents an approaching collection, represents the size of the adjacent set, represents the adjustment parameter used to control the sensitivity of local fitness; cluster analysis is performed through local fitness to classify similar digital microscopic images into one category, obtaining A microscopic imaging clusters, where A is an integer greater than zero. Commonly used clustering analysis methods include K-means clustering, hierarchical clustering, and density clustering; by performing feature annotation and cluster analysis on digital microscopic images of palisade cells of Cuscuta australis, different types of palisade cells can be more accurately identified and detected; feature annotation includes classification of microscopic observation morphology and target positioning, which enables the model to effectively distinguish cell types during the recognition process and improve detection accuracy.
[0046] Initialize the feature extraction model as a neural network model. The network layer structure includes: input layer, convolution layer, pooling layer, fully connected layer and output layer. Sigmoid function is used as activation function. The input layer uses the training set as input, and the output layer uses the feature cell map as output.
[0047] The feature recognition model is initialized as the Faster R-CNN model, and the region proposal network is defined with the Anchor mechanism as the candidate box generation mechanism. The classifier is the SoftMax classifier, and the regressor is the Lasso regressor. The feature cell map is used as the input of the feature recognition model, and the target recognition result is used as the output of the feature recognition model. The region proposal network and the Anchor mechanism are used to generate candidate boxes of different scales and aspect ratios, which enhances the model's ability to recognize targets in complex environments. Through the adjustment of the regressor and the discrimination of the classifier, the model can effectively splice candidate boxes of the same category, thereby improving the detection accuracy and recall rate.
[0048] The cross-training methods for the initial intelligent detection model of palisade cells of Cuscuta australis include:
[0049] Based on the total number K of clusters in the training cluster, K model trainings are performed. Each time the model is trained, one of the training clusters is used as the validation set, and the other training clusters are used as the training set for training, to obtain K initial intelligent detection models for palisade cells of southern dodder. The initial intelligent detection model for palisade cells of southern dodder includes a feature extraction model and a feature recognition model. For each model training, the digital microscopic imaging in the training set is used as the input of the feature extraction model. The digital microscopic imaging in the training set is convolved by the convolution layer to obtain a feature map. The feature map is pooled by the pooling layer to obtain a low-dimensional feature map. The low-dimensional feature map is feature mapped by the fully connected layer to obtain a feature cell map. The feature recognition model is used to perform feature recognition on the feature cell map to obtain a target recognition result. The confidence calculation is performed on the target recognition result. The formula for the confidence calculation is:
[0050] ;in Representative The confidence parameter of the target recognition result, Represents the total number of categories of identified content in the target recognition results, Represents the number of categories in the target recognition result that are consistent with the feature annotation content, Representative The target recognition results are the same as those of the The first feature annotation The correlation between categories, Represents the target recognition result The weight of each category, Representative The target recognition result categories, Representative The first feature annotation categories; the average value of the confidence parameters obtained in each training was calculated, and the initial southern dodder palisade cell intelligent detection model with the largest average value was taken as the optimal southern dodder palisade cell intelligent detection model.
[0051] The convolution operation on the digital microscopic images in the training set includes: presetting the convolution kernel's moving stride and the convolution matrix size, sliding the convolution kernel of the convolution layer in the digital microscopic image at the moving stride, and multiplying and summing the pixel elements of the selected digital microscopic image area with the same size as the convolution matrix by the convolution kernel after each slide to extract local features of the digital microscopic image;
[0052] The method of performing pooling operation on the feature map includes: dynamically adjusting the pooling window based on the size of the feature map and the size of the digital microscopy image. The length of the pooling window is calculated as follows:
[0053] ;in, represents the length of the pooling window, represents the length of digital microscopy imaging, represents the length of the feature map, Represents the floor function; the width of the pooling window is calculated as: ;in, Represents the width of the pooling window, represents the width of digital microscopy imaging, Represents the width of the feature map; select elements in the feature map according to the length and width of the pooling window. When the length of the pooling window is equal to the width, take the average value of the elements in the pooling window. When the length of the pooling window is not equal to the width, take the maximum value of the elements in the pooling window.
[0054] Methods for feature recognition of characteristic cell maps include:
[0055] The region proposal network scans the feature cell map and generates candidate boxes of different scales and aspect ratios using the anchor mechanism. The position and size of the bounding box are adjusted through the regressor. The classifier is used to identify the object and category of each candidate box. The candidate boxes identified as having the same category are then spliced together to obtain a new candidate box. Based on the size of the feature cell map and the size of the new candidate box, the area selected by the candidate box is converted into regional data. The converted regional data is the target recognition result.
[0056] The iterative optimization method for the optimal palisade cell intelligent detection model of Cuscuta australis includes:
[0057] A genetic algorithm is used to perform convolution optimization on the convolution layer in the feature extraction model to obtain the optimal convolution group. The optimal convolution group is used as the convolution kernel of the convolution layer in the feature extraction model to obtain the target feature extraction model. The digital microscopic imaging in the optimized cluster is used as the input of the target feature extraction model for feature extraction to obtain the characteristic cell maps corresponding to different optimized clusters. The cell map features corresponding to the same optimized cluster constitute a feature map cluster. The feature map cluster is used as the input of the feature recognition model for feature recognition to obtain the target recognition result. The target recognition result is trained by a comprehensive loss function to supervise the entire learning process. The non-maximum suppression algorithm is used to suppress the repeated results of different candidate boxes to obtain the target feature recognition model.
[0058] Ways to perform convolution optimization on the convolutional layers in the feature extraction model include:
[0059] For the convolutional layer in the feature extraction model, the default Initial convolution kernel groups, each convolution kernel group has E convolution kernels, initialize the optimal convolution group to be empty, initialize the weight parameters in each convolution kernel, and randomly take values based on the normal distribution L(a, b); based on the convolution kernel in the convolution layer, use the digital microscopy imaging in the optimized cluster as the input of the feature extraction model for forward propagation, use the output of the previous convolution layer as the input of the next convolution layer, use the cross entropy function as the loss function of each convolution layer, traverse the convolution kernel group, use each convolution kernel group as the convolution kernel in the convolution layer, and use the inverse of the loss function value as the convolution fitness of each convolution kernel group; calculate the selection probability of each convolution kernel group based on the convolution fitness, and the calculation formula of the selection probability is: ;in, Representative The selection probability of a convolution kernel group, Represents the index of the convolution kernel group, Representative The convolution fitness of the convolution kernel group, Representative The convolution fitness of the convolution kernel group is calculated; it is iteratively selected based on the selection probability, and F convolution kernel groups are selected as the parent convolution kernel groups each time. The parent convolution kernel groups are combined in pairs by the permutation and combination method to obtain the parent convolution pairs. For each pair of parent convolution pairs, the convolution kernel with the largest modulus value in the parent convolution kernel group is selected as the crossover node for single-point crossover to obtain the child convolution kernel group; the increment value is preset, and the increment value is a random small value to increase the diversity of the child convolution kernel group. The convolution kernels in the child convolution kernel group are traversed, and the weight parameters in the convolution kernel are selected and mutated by the random selection algorithm. Each step The weight parameters of the row selection mutation are added to the incremental value to obtain the mutated convolution kernel group; the mutated convolution kernel group is used as the convolution kernel in the convolution layer, and the convolution fitness of the mutated convolution kernel group is calculated. When the optimal convolution group is empty, the mutated convolution kernel group with the largest convolution fitness value is selected as the new optimal convolution group. When the optimal convolution group is not empty, the mutated convolution kernel group with the largest convolution fitness value that is greater than the convolution fitness value of the optimal convolution group is selected as the new optimal convolution group. Repeat until the convolution fitness of the optimal convolution group no longer changes, and output the optimal convolution group at this time as the convolution kernel of the convolution layer in the feature extraction model.
[0060] This embodiment utilizes artificial intelligence technologies such as deep learning to assist Chinese herbal medicine quality inspectors in conducting intelligent detection of palisade cells of Cuscuta australis and improve the accuracy and efficiency of screening work, and has significant application prospects. By combining deep learning technology with knowledge of Chinese herbal medicine identification and detection, a digital microscopic imaging-based intelligent detection technology for palisade cells of Cuscuta australis is realized, which further improves the accuracy and efficiency of Chinese herbal medicine identification and detection, effectively solves the defects in the detection process of palisade cells of Cuscuta australis, and promotes the application and development of a new generation of artificial intelligence technology in intelligent detection and auxiliary identification of Chinese herbal medicines. Through an efficient automated training process, the need for manual intervention is reduced, the level of intelligent detection is improved, and the needs of modern biological research and clinical applications are met.
[0061] In the examples of the present invention, a neural network model was trained using thousands of high-quality, high-resolution digital microscopic images of palisade cells of Cuscuta australis, acquired in collaboration with the Institute for the Authentication of Traditional Chinese and Ethnological Medicines. The annotation work in this invention was completed by several professional TCM material inspectors, who annotated over 1,200 digital microscopic images of palisade cells of Cuscuta australis.
[0062] Although preferred embodiments of the present invention have been described, those skilled in the art may make additional changes and modifications to these embodiments once they become aware of the basic inventive concepts. Therefore, the appended claims are intended to be interpreted as including the preferred embodiments and all changes and modifications that fall within the scope of the embodiments of the present invention.
[0063] Finally, it should be noted that, in this document, relational terms such as first and second, etc., are used only to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any actual relationship or order between these entities or operations. Similarly, the terms "comprises," "comprising," or any other variants thereof are intended to cover non-exclusive inclusion, so that a process, method, article, or terminal device that includes a series of elements includes not only those elements, but also other elements not explicitly listed, or elements that are inherent to such process, method, article, or terminal device. In the absence of further restrictions, an element defined by the sentence "comprising a ..." does not exclude the presence of other identical elements in the process, method, article, or terminal device that includes the element.
[0064] The above is a detailed introduction to the training method and device for the intelligent detection model of palisade cells of southern dodder provided by this application. Specific examples are used in this article to illustrate the principles and implementation methods of this application. The description of the above embodiments is only used to help understand the method of this application and its core idea; at the same time, for general technical personnel in this field, based on the ideas of this application, there will be changes in the specific implementation methods and application scope. In summary, the content of this specification should not be understood as a limitation on this application.
Claims
1. A method for training an intelligent detection model for palisade cells of Cuscuta australis, characterized in that: include: S1. Collect digital microscopic images of palisade cells of Cuscuta australis, with each digital microscopic image labeled with features; based on Feature annotation performs random aggregation on digital microscopic images to obtain A microscopic imaging clusters, which are then split into clusters to obtain training clusters and optimization clusters. S2. Build an initial intelligent detection model for palisade cells of Cuscuta australis based on deep learning technology; S3, cross-training the initial southern dodder palisade cell intelligent detection model using the training cluster to obtain the optimal southern dodder palisade cell intelligent detection model; S4. Iteratively optimize the optimal southern dodder palisade cell intelligent detection model using the optimization cluster to obtain the target southern dodder palisade cell intelligent detection model; The method of randomly aggregating digital microscopic imaging includes: Construct microscopic feature vectors based on feature annotation, use similarity calculation formula to calculate the similarity between different microscopic feature vectors, and construct similarity matrix based on similarity ;by Represents the similarity matrix Rank The elements of the column, Representative The microscopic eigenvectors and The similarity between the microscopic eigenvectors; for the similarity matrix The index of the column whose element value is greater than the preset similarity threshold constitutes the first The adjacent set of microscopic feature vectors is constructed; the fitness of each microscopic feature vector is calculated based on the similarity matrix and the adjacent set. The fitness calculation formula is: ;in, Representative The local fitness of the microscopic eigenvectors, Represents an approaching collection, represents the size of the adjacent set, represents the adjustment parameter; cluster analysis is performed through local fitness, similar digital microscopy images are classified into one category, and A microscopy clusters are obtained, where A is an integer greater than zero.
2. The method for training an intelligent detection model of palisade cells of Cuscuta australis according to claim 1, characterized in that: The feature annotation includes: feature classification annotation and target positioning annotation; target positioning annotation includes: marking position coordinates on digital microscopic imaging; feature classification annotation includes: marking image features of palisade cells of southern dodder on digital microscopic imaging.
3. The method for training an intelligent detection model of palisade cells of Cuscuta australis according to claim 2, characterized in that: The construction method of the initial intelligent detection model of the palisade cells of Cuscuta australis includes: Initialize the feature extraction model as a neural network model. The network layer structure includes: input layer, convolution layer, pooling layer, fully connected layer and output layer. Sigmoid function is used as activation function. The input layer uses the training set as input, and the output layer uses the feature cell map as output. The feature recognition model is initialized as the Faster R-CNN model, and the region proposal network is defined to use the Anchor mechanism as the candidate box generation mechanism. The classifier is the SoftMax classifier, the regressor is the Lasso regressor, the feature cell map is used as the input of the feature recognition model, and the target recognition result is used as the output of the feature recognition model.
4. The method for training an intelligent detection model for palisade cells of Cuscuta australis according to claim 3, wherein: The method of cross-training the initial Cuscuta australis palisade cell intelligent detection model includes: Based on the total number K of clusters in the training cluster, K model trainings are performed. Each time the model training is performed, one of the training clusters is used as the validation set, and the other training clusters are used as the training set for training, to obtain K initial southern dodder palisade cell intelligent detection models; the initial southern dodder palisade cell intelligent detection model includes a feature extraction model and a feature recognition model. For each model training, the digital microscopic imaging in the training set is used as the input of the feature extraction model, and the digital microscopic imaging in the training set is convolved through the convolution layer to obtain a feature map; the feature map is pooled through the pooling layer to obtain a low-dimensional feature map; the low-dimensional feature map is feature mapped through the fully connected layer to obtain a feature cell map; the feature recognition model is used to perform feature recognition on the feature cell map to obtain a target recognition result, and the target recognition result is confidence calculated to obtain a confidence parameter; the average value of the confidence parameters obtained from each training is calculated, and the initial southern dodder palisade cell intelligent detection model with the largest average value is used as the optimal southern dodder palisade cell intelligent detection model.
5. The method for training an intelligent detection model of palisade cells of Cuscuta australis according to claim 4, characterized in that: The method of performing a convolution operation on the digital microscopic images in the training set includes: presetting a convolution kernel movement stride and a convolution matrix size, sliding the convolution kernel of the convolution layer in the digital microscopic images at the movement stride, and multiplying and summing pixel elements of a selected digital microscopic imaging area having the same size as the convolution matrix by the convolution kernel after each slide, thereby extracting local features of the digital microscopic images; The method of performing the pooling operation on the feature map includes: dynamically adjusting the pooling window based on the size of the feature map and the size of the digital microscopic image, and the length of the pooling window is calculated as follows: ;in, represents the length of the pooling window, represents the length of digital microscopy imaging, represents the length of the feature map, Represents the floor function; the width of the pooling window is calculated as: ;in, Represents the width of the pooling window, represents the width of digital microscopy imaging, Represents the width of the feature map; select elements in the feature map according to the length and width of the pooling window. When the length of the pooling window is equal to the width, take the average value of the elements in the pooling window. When the length of the pooling window is not equal to the width, take the maximum value of the elements in the pooling window.
6. The method for training an intelligent detection model for palisade cells of Cuscuta australis according to claim 5, characterized in that: The method of performing feature recognition on the characteristic cell map includes: The region proposal network scans the feature cell map and generates candidate boxes of different scales and aspect ratios using the anchor mechanism. The position and size of the bounding box are adjusted through the regressor. The classifier is used to identify the object and category of each candidate box. The candidate boxes identified as having the same category are then spliced together to obtain a new candidate box. Based on the size of the feature cell map and the size of the new candidate box, the area selected by the candidate box is converted into regional data. The converted regional data is the target recognition result.
7. The method for training an intelligent detection model for palisade cells of Cuscuta australis according to claim 6, characterized in that: The formula for calculating the confidence of the target recognition result is: ;in Representative The confidence parameter of the target recognition result, Represents the total number of categories of identified content in the target recognition results, Represents the number of categories in the target recognition result that are consistent with the feature annotation content, Representative The target recognition results are the same as those of the The first feature annotation The correlation between categories, Represents the target recognition result The weight of each category, Representative The target recognition result categories, Representative The first feature annotation categories.
8. The method for training an intelligent detection model for palisade cells of Cuscuta australis according to claim 7, characterized in that: The method of iteratively optimizing the optimal palisade cell intelligent detection model for Cuscuta australis comprises: A genetic algorithm is used to perform convolution optimization on the convolution layer in the feature extraction model to obtain the optimal convolution group. The optimal convolution group is used as the convolution kernel of the convolution layer in the feature extraction model to obtain the target feature extraction model. The digital microscopic imaging in the optimized cluster is used as the input of the target feature extraction model for feature extraction to obtain characteristic cell maps corresponding to different optimized clusters. The cell map features corresponding to the same optimized cluster constitute a feature map cluster. The feature map cluster is used as the input of the feature recognition model for feature recognition to obtain target recognition results. The target recognition results are trained using a comprehensive loss function. The non-maximum suppression algorithm is used to suppress the repeated results of different candidate boxes to obtain the target feature recognition model.
9. The method for training an intelligent detection model for palisade cells of Cuscuta australis according to claim 8, characterized in that: The method of performing convolution optimization on the convolution layer in the feature extraction model includes: For the convolutional layer in the feature extraction model, the default Initial convolution kernel groups, each convolution kernel group has E convolution kernels, initialize the optimal convolution group to be empty, initialize the weight parameters in each convolution kernel, and randomly take values based on the normal distribution L(a, b); based on the convolution kernel in the convolution layer, use the digital microscopy imaging in the optimized cluster as the input of the feature extraction model for forward propagation, use the output of the previous convolution layer as the input of the next convolution layer, use the cross entropy function as the loss function of each convolution layer, traverse the convolution kernel group, use each convolution kernel group as the convolution kernel in the convolution layer, and use the inverse of the loss function value as the convolution fitness of each convolution kernel group; calculate the selection probability of each convolution kernel group based on the convolution fitness, and the calculation formula of the selection probability is: ;in, Representative The selection probability of a convolution kernel group, Represents the index of the convolution kernel group, Representative The convolution fitness of the convolution kernel group, Representative The convolution fitness of the convolution kernel group; iterative selection is performed based on the selection probability, and F convolution kernel groups are selected as the parent convolution kernel groups each time. The parent convolution kernel groups are combined in pairs using the permutation and combination method to obtain the parent convolution pairs. For each pair of parent convolution pairs, the convolution kernel with the largest modulus value in the parent convolution kernel group is selected as the crossover node for single-point crossover to obtain the child convolution kernel group; the increment value is preset, the convolution kernels in the child convolution kernel group are traversed, and the weight parameters in the convolution kernel are selected and mutated using the random selection algorithm. Each weight parameter selected and mutated is consistent with the increment value. The values are added to obtain the mutated convolution kernel group; the mutated convolution kernel group is used as the convolution kernel in the convolution layer, and the convolution fitness of the mutated convolution kernel group is calculated. When the optimal convolution group is empty, the mutated convolution kernel group with the largest convolution fitness value is selected as the new optimal convolution group. When the optimal convolution group is not empty, the mutated convolution kernel group with the largest convolution fitness value that is greater than the convolution fitness value of the optimal convolution group is selected as the new optimal convolution group. Repeat until the convolution fitness of the optimal convolution group no longer changes, and output the optimal convolution group at this time as the convolution kernel of the convolution layer in the feature extraction model.