TDR-Model: Tomato Disease Recognition Based on Image Dehazing and Improved MobileNetV3 Model
Currently, deep learning applications in tomato disease recognition are becoming widespread. However, in greenhouse tomato cultivation environments, many techniques neglect the foggy image caused by wet conditions, adversely affecting the performance of disease recognition models, which still remain...
Saved in:
| Main Authors: | , , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
IEEE
2025-01-01
|
| Series: | IEEE Access |
| Subjects: | |
| Online Access: | https://ieeexplore.ieee.org/document/10813355/ |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1841563324260548608 |
|---|---|
| author | Zhixiang Zhang Tong Liu Jinyu Gao Meng Yang Wenjun Luo Fanqiang Lin |
| author_facet | Zhixiang Zhang Tong Liu Jinyu Gao Meng Yang Wenjun Luo Fanqiang Lin |
| author_sort | Zhixiang Zhang |
| collection | DOAJ |
| description | Currently, deep learning applications in tomato disease recognition are becoming widespread. However, in greenhouse tomato cultivation environments, many techniques neglect the foggy image caused by wet conditions, adversely affecting the performance of disease recognition models, which still remains unresolved. To alleviate adverse environmental impacts, such as fog, and enhance the precision and robustness of the model, this study introduces an improved dark channel prior (DCP) dehazing algorithm and proposes the TDR-Model based on the MobileNetV3. Employing the quad-tree spatial hierarchy search strategy, guided filtering, tolerance mechansim, and multi-scale color restoration (MSRCR) algorithm, an innovative improved version of DCP dehazing algorithm is realized. Image preprocessing with the improved DCP dehazing algorithm obtains more authentic dehazed images. Moreover, MobileNetV3 is improved by the incorporation of the convolutional block attention module (CBAM) and Omni-Dimensional Dynamic Convolution(ODC), improving the accuracy of disease feature recognition. The experiment results demonstrate that, with the parameter size remaining nearly unchanged (a slight reduction of 0.04M), an increase of 16.88% is achieved in accuracy compared to the original version for foggy images. Additionally, we provide extensive ablation studies to evaluate effectiveness of the proposed framework. |
| format | Article |
| id | doaj-art-d29cf168e477489b8feba109ae6bef15 |
| institution | Kabale University |
| issn | 2169-3536 |
| language | English |
| publishDate | 2025-01-01 |
| publisher | IEEE |
| record_format | Article |
| series | IEEE Access |
| spelling | doaj-art-d29cf168e477489b8feba109ae6bef152025-01-03T00:02:02ZengIEEEIEEE Access2169-35362025-01-011385286510.1109/ACCESS.2024.352210110813355TDR-Model: Tomato Disease Recognition Based on Image Dehazing and Improved MobileNetV3 ModelZhixiang Zhang0https://orcid.org/0009-0000-6394-7646Tong Liu1https://orcid.org/0009-0008-5046-8417Jinyu Gao2https://orcid.org/0009-0009-6968-9508Meng Yang3https://orcid.org/0009-0009-7824-7306Wenjun Luo4Fanqiang Lin5https://orcid.org/0000-0003-4884-1944College of Mechanical and Electrical Engineering, Chengdu University of Technology, Chengdu, ChinaCollege of Mechanical and Electrical Engineering, Chengdu University of Technology, Chengdu, ChinaCollege of Mechanical and Electrical Engineering, Chengdu University of Technology, Chengdu, ChinaCollege of Mechanical and Electrical Engineering, Chengdu University of Technology, Chengdu, ChinaSantai County Education and Sports Bureau, Mianyang, ChinaCollege of Mechanical and Electrical Engineering, Chengdu University of Technology, Chengdu, ChinaCurrently, deep learning applications in tomato disease recognition are becoming widespread. However, in greenhouse tomato cultivation environments, many techniques neglect the foggy image caused by wet conditions, adversely affecting the performance of disease recognition models, which still remains unresolved. To alleviate adverse environmental impacts, such as fog, and enhance the precision and robustness of the model, this study introduces an improved dark channel prior (DCP) dehazing algorithm and proposes the TDR-Model based on the MobileNetV3. Employing the quad-tree spatial hierarchy search strategy, guided filtering, tolerance mechansim, and multi-scale color restoration (MSRCR) algorithm, an innovative improved version of DCP dehazing algorithm is realized. Image preprocessing with the improved DCP dehazing algorithm obtains more authentic dehazed images. Moreover, MobileNetV3 is improved by the incorporation of the convolutional block attention module (CBAM) and Omni-Dimensional Dynamic Convolution(ODC), improving the accuracy of disease feature recognition. The experiment results demonstrate that, with the parameter size remaining nearly unchanged (a slight reduction of 0.04M), an increase of 16.88% is achieved in accuracy compared to the original version for foggy images. Additionally, we provide extensive ablation studies to evaluate effectiveness of the proposed framework.https://ieeexplore.ieee.org/document/10813355/Tomato disease recognitiondark channel prior dehazing algorithmdeep learningimage preprocessinglightweight model |
| spellingShingle | Zhixiang Zhang Tong Liu Jinyu Gao Meng Yang Wenjun Luo Fanqiang Lin TDR-Model: Tomato Disease Recognition Based on Image Dehazing and Improved MobileNetV3 Model IEEE Access Tomato disease recognition dark channel prior dehazing algorithm deep learning image preprocessing lightweight model |
| title | TDR-Model: Tomato Disease Recognition Based on Image Dehazing and Improved MobileNetV3 Model |
| title_full | TDR-Model: Tomato Disease Recognition Based on Image Dehazing and Improved MobileNetV3 Model |
| title_fullStr | TDR-Model: Tomato Disease Recognition Based on Image Dehazing and Improved MobileNetV3 Model |
| title_full_unstemmed | TDR-Model: Tomato Disease Recognition Based on Image Dehazing and Improved MobileNetV3 Model |
| title_short | TDR-Model: Tomato Disease Recognition Based on Image Dehazing and Improved MobileNetV3 Model |
| title_sort | tdr model tomato disease recognition based on image dehazing and improved mobilenetv3 model |
| topic | Tomato disease recognition dark channel prior dehazing algorithm deep learning image preprocessing lightweight model |
| url | https://ieeexplore.ieee.org/document/10813355/ |
| work_keys_str_mv | AT zhixiangzhang tdrmodeltomatodiseaserecognitionbasedonimagedehazingandimprovedmobilenetv3model AT tongliu tdrmodeltomatodiseaserecognitionbasedonimagedehazingandimprovedmobilenetv3model AT jinyugao tdrmodeltomatodiseaserecognitionbasedonimagedehazingandimprovedmobilenetv3model AT mengyang tdrmodeltomatodiseaserecognitionbasedonimagedehazingandimprovedmobilenetv3model AT wenjunluo tdrmodeltomatodiseaserecognitionbasedonimagedehazingandimprovedmobilenetv3model AT fanqianglin tdrmodeltomatodiseaserecognitionbasedonimagedehazingandimprovedmobilenetv3model |