Diagnosis of Custard Apple Disease Based on Adaptive Information Entropy Data Augmentation and Multiscale Region Aggregation Interactive Visual Transformers
Accurate diagnosis of plant diseases is crucial for crop health. This study introduces the EDA–ViT model, a Vision Transformer (ViT)-based approach that integrates adaptive entropy-based data augmentation for diagnosing custard apple <i>(Annona squamosa</i>) diseases. Traditional models...
Saved in:
| Main Authors: | , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
MDPI AG
2024-11-01
|
| Series: | Agronomy |
| Subjects: | |
| Online Access: | https://www.mdpi.com/2073-4395/14/11/2605 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1850149568851738624 |
|---|---|
| author | Kunpeng Cui Jianbo Huang Guowei Dai Jingchao Fan Christine Dewi |
| author_facet | Kunpeng Cui Jianbo Huang Guowei Dai Jingchao Fan Christine Dewi |
| author_sort | Kunpeng Cui |
| collection | DOAJ |
| description | Accurate diagnosis of plant diseases is crucial for crop health. This study introduces the EDA–ViT model, a Vision Transformer (ViT)-based approach that integrates adaptive entropy-based data augmentation for diagnosing custard apple <i>(Annona squamosa</i>) diseases. Traditional models like convolutional neural network and ViT face challenges with local feature extraction and large dataset requirements. EDA–ViT overcomes these by using a multi-scale weighted feature aggregation and a feature interaction module, enhancing both local and global feature extraction. The adaptive data augmentation method refines the training process, boosting accuracy and robustness. With a dataset of 8226 images, EDA–ViT achieved a classification accuracy of 96.58%, an F1 score of 96.10%, and a Matthews Correlation Coefficient (MCC) of 92.24%, outperforming other models. The inclusion of the Deformable Multi-head Self-Attention (DMSA) mechanism further enhanced feature capture. Ablation studies revealed that the adaptive augmentation contributed to a 0.56% accuracy improvement and a 0.34% increase in MCC. In summary, EDA–ViT presents an innovative solution for custard apple disease diagnosis, with potential applications in broader agricultural disease detection, ultimately aiding precision agriculture and crop health management. |
| format | Article |
| id | doaj-art-db7949c61cf547c9a024bf5e9e4a8b8c |
| institution | OA Journals |
| issn | 2073-4395 |
| language | English |
| publishDate | 2024-11-01 |
| publisher | MDPI AG |
| record_format | Article |
| series | Agronomy |
| spelling | doaj-art-db7949c61cf547c9a024bf5e9e4a8b8c2025-08-20T02:26:51ZengMDPI AGAgronomy2073-43952024-11-011411260510.3390/agronomy14112605Diagnosis of Custard Apple Disease Based on Adaptive Information Entropy Data Augmentation and Multiscale Region Aggregation Interactive Visual TransformersKunpeng Cui0Jianbo Huang1Guowei Dai2Jingchao Fan3Christine Dewi4Ji Yang College, Zhejiang A&F University, Shaoxing 311800, ChinaJi Yang College, Zhejiang A&F University, Shaoxing 311800, ChinaCollege of Computer Science, Sichuan University, Chengdu 610065, ChinaAgricultural Information Institute of CAAS, National Agriculture Science Data Center, Beijing 100081, ChinaSchool of Information Technology, Deakin University, Campus, 221 Burwood Hwy, Burwood, VIC 3125, AustraliaAccurate diagnosis of plant diseases is crucial for crop health. This study introduces the EDA–ViT model, a Vision Transformer (ViT)-based approach that integrates adaptive entropy-based data augmentation for diagnosing custard apple <i>(Annona squamosa</i>) diseases. Traditional models like convolutional neural network and ViT face challenges with local feature extraction and large dataset requirements. EDA–ViT overcomes these by using a multi-scale weighted feature aggregation and a feature interaction module, enhancing both local and global feature extraction. The adaptive data augmentation method refines the training process, boosting accuracy and robustness. With a dataset of 8226 images, EDA–ViT achieved a classification accuracy of 96.58%, an F1 score of 96.10%, and a Matthews Correlation Coefficient (MCC) of 92.24%, outperforming other models. The inclusion of the Deformable Multi-head Self-Attention (DMSA) mechanism further enhanced feature capture. Ablation studies revealed that the adaptive augmentation contributed to a 0.56% accuracy improvement and a 0.34% increase in MCC. In summary, EDA–ViT presents an innovative solution for custard apple disease diagnosis, with potential applications in broader agricultural disease detection, ultimately aiding precision agriculture and crop health management.https://www.mdpi.com/2073-4395/14/11/2605plant diseaseconvolutional neural networkadaptive data augmentationfeature fusionvisual transformer |
| spellingShingle | Kunpeng Cui Jianbo Huang Guowei Dai Jingchao Fan Christine Dewi Diagnosis of Custard Apple Disease Based on Adaptive Information Entropy Data Augmentation and Multiscale Region Aggregation Interactive Visual Transformers Agronomy plant disease convolutional neural network adaptive data augmentation feature fusion visual transformer |
| title | Diagnosis of Custard Apple Disease Based on Adaptive Information Entropy Data Augmentation and Multiscale Region Aggregation Interactive Visual Transformers |
| title_full | Diagnosis of Custard Apple Disease Based on Adaptive Information Entropy Data Augmentation and Multiscale Region Aggregation Interactive Visual Transformers |
| title_fullStr | Diagnosis of Custard Apple Disease Based on Adaptive Information Entropy Data Augmentation and Multiscale Region Aggregation Interactive Visual Transformers |
| title_full_unstemmed | Diagnosis of Custard Apple Disease Based on Adaptive Information Entropy Data Augmentation and Multiscale Region Aggregation Interactive Visual Transformers |
| title_short | Diagnosis of Custard Apple Disease Based on Adaptive Information Entropy Data Augmentation and Multiscale Region Aggregation Interactive Visual Transformers |
| title_sort | diagnosis of custard apple disease based on adaptive information entropy data augmentation and multiscale region aggregation interactive visual transformers |
| topic | plant disease convolutional neural network adaptive data augmentation feature fusion visual transformer |
| url | https://www.mdpi.com/2073-4395/14/11/2605 |
| work_keys_str_mv | AT kunpengcui diagnosisofcustardapplediseasebasedonadaptiveinformationentropydataaugmentationandmultiscaleregionaggregationinteractivevisualtransformers AT jianbohuang diagnosisofcustardapplediseasebasedonadaptiveinformationentropydataaugmentationandmultiscaleregionaggregationinteractivevisualtransformers AT guoweidai diagnosisofcustardapplediseasebasedonadaptiveinformationentropydataaugmentationandmultiscaleregionaggregationinteractivevisualtransformers AT jingchaofan diagnosisofcustardapplediseasebasedonadaptiveinformationentropydataaugmentationandmultiscaleregionaggregationinteractivevisualtransformers AT christinedewi diagnosisofcustardapplediseasebasedonadaptiveinformationentropydataaugmentationandmultiscaleregionaggregationinteractivevisualtransformers |