Project Articles

Default Latest Most Read

Please wait a minute... For Selected: Download Citations EndNote Ris BibTeX Toggle Thumbnails Select Identification of Citrus Diseases Based on Improved ShuffleNet V2 YU Yannan, MO Yongbin, YAN Jichi, XIONG Chunlin, DOU Shiqing, YANG Rongfeng Journal of Henan Agricultural Sciences    2024, 53 (1): 142-151.   DOI: 10.15933/j.cnki.1004-3268.2024.01.016 Abstract （766）      PDF （3756KB）（54）       Knowledge map    Save Large convolutional neural networks are difficult to deploy in practical applications because of the complexity of models，while lightweight networks are often less accurate than the former because of the optimization of model structure. To solve these problems，ShuffleNet V2 was improved and a lightweight MAM‑ShuffleNet citrus disease recognition model was proposed. Firstly，the mixed attention module（MAM）was introduced in ShuffleNet V2 to improve the ability of the model to extract disease features. Secondly，Ghost module was used to optimize the convolutional layer in the network，which effectively reduced the number of network model parameters and calculation cost. Finally，the stacking times of ShuffleNet V2 unit in the network structure were adjusted to further simplify the network parameters. The results showed that the average recognition accuracy of MAM‑ShuffleNet model reached 97.7% in the self‑built citrus leaf data. Compared with the original ShuffleNet V2，the number of parameters was reduced by 45.7%，and the recognition accuracy was increased by 1.2 percentage points.The comprehensive performance was better than ResNet50 and DenseNet121 models. Reference | Related Articles | Metrics Select Research on Millet Disease Identification Based on Transfer Learning and Residual Network ZHANG Hongtao, LUO Yiming, TAN Lian, YANG Jiapeng, WANG Yu Journal of Henan Agricultural Sciences    2023, 52 (12): 162-171.   DOI: 10.15933/j.cnki.1004-3268.2023.12.018 Abstract （742）      PDF （2368KB）（55）       Knowledge map    Save A method of millet disease image recognition based on transfer learning and residual network（Residual CNN）was proposed for millet disease. First，the original sample set was established，whichwas composed of four kinds of disease images including millet white disease，blast，red leaf disease，rustdisease and normal millet leaf image.Then，the original image was segmented by using the maximum inter‐class variance method based on super green feature，the millet disease segmentation image dataset was established，and the dataset was extended. Finally，based on the expanded segmentation image data set of millet disease，the recognition model of millet disease was established by using the idea of transfer learning and residual network. The results showed that the recognition rate of this model reached 98.2%，which was 8.9 percentage points higher than that of the support vector machine（SVM）based millet disease recognition model，and the training time of this model was reduced by 17.69 min compared with that of the convolutional neural network（CNN） based millet disease recognition model.The results indicated that the recognition model of millet disease based on transfer learning and residual network could effectively identify the four kinds of millet leaf diseases. Reference | Related Articles | Metrics Select Identification of Lasioderma serricorne in Tobacco Leaf Raw Materials Based on Improved YOLOv3 Algorithm ZHANG Weizheng, CHEN Saiyueyang, WANG Yanling, TIE Jinxin, DING Jia, LI Meng, LI Canlin, SU Xiaoke, GAN Yong Journal of Henan Agricultural Sciences    2023, 52 (11): 157-166.   DOI: 10.15933/j.cnki.1004-3268.2023.11.017 Abstract （679）      PDF （10647KB）（92）       Knowledge map    Save In order to solve the problem of low efficiency and inaccuracy in manual recognition of tobacco beetles（ Lasioderma serricorne）in tobacco leaf raw materials，and achieve accurate and intelligent recognition of tobacco beetles，a tobacco beetle recognition model was proposed based on the improved YOLOv3 algorithm.Firstly，Random Mix data augmentation technology was used to expand the tobacco beetle dataset；By introducing the K⁃means++ algorithm to recluster anchor boxes，YOLOv3’s recognition ability for tobacco beetles was improved；SIoU Loss was used to improve the Loss function of YOLOv3 bounding box to improve the accuracy of model positioning and accelerate model convergence；Finally，a feature refinement module was added to filter conflict information and optimize the accuracy of the model in identifying small targets in the tobacco beetle dataset.The results showed that，the improved YOLOv3 network model achieved average detection accuracy，recall，F1 score，and mAP of 93.26%，88.25%，0.90，and 94.59% for tobacco beetles in the test set，respectively，increased by 12.21 percentage points，11.79 percentage points，0.12，and 12.40 percentage points compared to the existing YOLOv3 algorithm，providing support for the effective recognition of tobacco beetles. Reference | Related Articles | Metrics Select Apple Leaf Diseases Identification Based on Improved Residual Network CHEN Cong, YU Xiao, GONG Qi Journal of Henan Agricultural Sciences    2023, 52 (4): 152-161.   DOI: 10.15933/j.cnki.1004-3268.2023.04.018 Abstract （984）      PDF （5367KB）（215）       Knowledge map    Save Apple leaf diseases are similar in morphology and different in spot size，and traditional methods relying on labor and agricultural experts to identify them are inefficient. In view of this，this study proposed an apple disease identification model REP⁃ResNet based on improved residual network.This model was improved by adopting the order adjustment of residual structure（that was bath normalization，activation function，convolutional layer），and adding the efficient channel attention and parallel convolution based on the baseline model ResNet⁃50.During the training process，the model weight parameters pretrained in the public dataset PlantVillage were transferred to the above network model for retraining，which aimed to accelerate the convergence speed of the network and improve the recognition ability of the model.Data expansion was used to solve the problem of uneven samples during training.The results showed that the recognition accuracy of the REP⁃ResNet model was 2.41 percentage points higher than that of the benchmark network model.The model was trained by transfer learning，and the accuracy rate of apple leaf disease recognition in complex backgrounds reached 97. 69%. Compared with traditional convolutional neural networks，the recognition effect was greatly improved. Reference | Related Articles | Metrics Select Identification of Crop Leaf Diseases Based on Improved MobileNetV2 Model WANG Huanxin, SHEN Zhihao, LIU Quan, LIU Jinjiang Journal of Henan Agricultural Sciences    2023, 52 (4): 143-151.   DOI: 10.15933/j.cnki.1004-3268.2023.04.017 Abstract （1058）      PDF （2715KB）（196）       Knowledge map    Save In order to achieve convenient recognition of crop leaf disease images based on mobile，improve the efficiency of crop disease recognition and thus better guide crop disease control，a crop disease recognition model was established based on an improved lightweight convolutional neural network MobileNetV2.First，a crop dataset containing 15 types of diseased leaves and 4 types of healthy leaves was created and the data was balanced using data enhancement operations.Second，based on the improvement of MobileNetV2，efficient channel attention（ECA）and attentional feature fusion（AFF）were introduced，and the redundant layers were removed by model pruning，and a high⁃performance lightweight crop disease recognition model was proposed.The results showed that the number of parameters of the improved MobileNetV2 model was reduced by 15.37% compared with that of MobileNetV2，while the recognition accuracy was improved by 0.9 percentage points to 98.4% compared with that before the improvement.Compared with the classical convolutional neural network models such as EfficientNet⁃b0 and ShuffleNetV2⁃0. 5X，the improved model not only had the highest recognition accuracy，but also had a faster convergence rate during training. Reference | Related Articles | Metrics Select Research on Maize Pest Detection Based on Instance Segmentation ZHAO Kangdi, SHAN Yugang, YUAN Jie, ZHAO Yuanlong Journal of Henan Agricultural Sciences    2022, 51 (12): 153-161.   DOI: 10.15933/j.cnki.1004-3268.2022.12.018 Abstract （590）      PDF （5388KB）（109）       Knowledge map    Save In order to achieve accurate and rapid identification of maize pests，this paper proposes a pest detection method using convolutional neural network combined with transfer learning based on instance segmentation.Taking eggs，larvae and adults of Spodoptera frugiperda as detection objects，the image data was expanded by data enhancement，and the pre‑training weights of Yolact++ model on COCO data set were migrated to the detection of Spodoptera frugiperda.Using the idea of focal loss to solve the imbalance of difficult and easy samples，the loss function in the model was optimized. The Detnet model was used to improve the Resnet trunk model in the Yolact++ model to improve the effect of small target detection.In the training process，the convolution layer was first frozen and then thawed，and the local and global training methods were combined to obtain the optimal weight model，and the model was tested. The test results showed that compared with the detection algorithms of YoloV3+migration learning，SSD+migration learning，Yolact+migration learning and Yolact++，this method had better accuracy and missed detection rate for complex background image detection. The accuracy of the test reached 96.32%，the missed detection rate was 5.51%，and the false detection rate was 5.33%. Reference | Related Articles | Metrics Select Research on Identification Accurate Rate of Soybean Leaf Diseases Based on UAV Image Processing TAN Qinhong Journal of Henan Agricultural Sciences    2021, 50 (3): 174-180.   DOI: 10.15933/j.cnki.1004-3268.2021.03.023 Abstract （940）      PDF （2997KB）（251）       Knowledge map    Save In order to balance the contradiction between unmanned aerial vehicle(UAV) recognition coverage and recognition accuracy, and realize the identification accurate rate of large-area soybean diseases,SLIC super-pixel method was used to process soybean leaf images acquired by UAV at 1 m、2 m、4 m、8 m、16 m.After extracting the color,gradient,texture and shape characteristics of leaves,sequential minimal optimization,J48 decision tree,k-nearest neighbors and random forest algorithm were used based on the feature vectors of these attributes to identify soybean leaf diseases,and the accuracy of different algorithms for disease identification was obtained.The influence of input parameters on disease identification was analyzed by comparing the accuracy of disease recognition with feature fusion and single feature.The results showed that the accurate rates of the four algorithms were more than 90% when the UAV image shooting height was 1 m and 2 m,and the SMO algorithm and random forest algorithm had higher identification accurate rate.The identification accurate rate of soybean disease with fusion feature vector as input parameter was higher than that of single feature vector. Color was the attribute that could best reflect the actual situation of leaf disease.The optimal shooting height of UAV was 1—2 m. Related Articles | Metrics