Academic Journal of Computing & Information Science, 2024, 7(2); doi: 10.25236/AJCIS.2024.070210.
Yueyang Wu1, Xinyao Zhang2, Minhua Ye3, Guiyuan Zhang1, Xiang Zhang1, Yize Tang1, Xuewen Chen1, Yongpei Cao1, Derong Yang1, Zhi Li1
1School of Mathematics and Computer, Guangdong Ocean University, Zhanjiang, 524088, China
2College of Food Science and Technology, Guangdong Ocean University, Zhanjiang, 524088, China
3College of Ocean Engineering and Energy, Guangdong Ocean University, Zhanjiang, 524088, China
To address the issue of low accuracy and poor performance of existing crop disease detection methods in identifying various tomato leaf diseases, this study proposes an improved tomato leaf disease detection model, TMT-YOLOv5s, based on the YOLOv5 network model. Initially, a DCAM attention mechanism module is constructed within the Backbone of the original YOLOv5s model. This module enhances the model's capability to extract pathological features of tomato leaves by implementing dual-channel attention and spatial attention mechanisms and integrates with the BiFPN module to mitigate the influence of complex background features. The aim is to improve the model's accuracy and precision in detecting and classifying different types of diseases. Experimental results demonstrate that the TMT-YOLOv5s model achieves an average precision of 98.59% and a recall rate of 96.31%, marking an improvement of 2.79 percentage points and 2.51 percentage points, respectively, compared to the original YOLOv5s model. The model exhibits superior detection accuracy and effectiveness, accurately identifying various types of tomato leaf diseases. These findings provide valuable insights for the practical application of tomato leaf disease detection.
Deep learning, YOLOv5s, Convolutional Neural Network, Disease detection, Tomato leaf diseases
Yueyang Wu, Xinyao Zhang, Minhua Ye, Guiyuan Zhang, Xiang Zhang, Yize Tang, Xuewen Chen, Yongpei Cao, Derong Yang, Zhi Li. A Lightweight Tomato Leaf Disease Detection System with Strong Generalization Ability Based on TMT-YOLOv5s. Academic Journal of Computing & Information Science (2024), Vol. 7, Issue 2: 72-78. https://doi.org/10.25236/AJCIS.2024.070210.
[1] Abeid E E S ,Mosa A M ,Tabakh E M A M , et al.Antifungal activity of copper oxide nanoparticles derived from Zizyphus spina leaf extract against Fusarium root rot disease in tomato plants.[J]. Journal of Nanobiotechnology, 2024, 22(1):28-28.
[2] Ruihan C, Minhua Y, Zhi L, et al. Empirical assessment of carbon emissions in Guangdong Province within the framework of carbon peaking and carbon neutrality: a lasso-TPE-BP neural network approach.[J]. Environmental science and pollution research international, 2023, 30(58):121647-121665.
[3] Wen Y, Chen Y, Wang K, et al. A Review of Pest Detection Based on Machine Vision[J]. Journal of China Oils and Grains Society, 2022, 37(10): 271-279.
[4] Liu J, Wang X. Tomato Disease and Pest Recognition Algorithm Based on YOLO[J]. China Vegetables, 2020, 33(9): 18-22, 38.
[5] Wang W, Liu Z, Gao P, et al. A Lychee Disease and Pest Detection Model Based on Improved YOLOv4 [J]. Transactions of the Chinese Society for Agricultural Machinery, 2023, 54(5): 227-235.
[6] Liu G, Hu G, Gulibahaertuoheti, et al. Detection Method of Grape Leaf Diseases and Pests Based on Improved YOLOv3[J]. Microelectronics & Computer, 2023, 40(2): 110-119.
[7] Cai P ,Yang D ,Zou Y , et al.Reconsidering Multi-Branch Aggregation for Semantic Segmentation[J]. Electronics, 2023, 12(15).
[8] Luo J ,Wang J ,Wen B , et al.Accelerated genetic algorithm based on real number coding for solving mobile communication network site planning[J].Academic Journal of Computing Information Science, 2023, 6(7).