In image classification of tomato leaf diseases based on deep learning, models often focus on features such as edges, stems, backgrounds, and shadows of the experimental samples, while ignoring the features of the disease area, resulting in weak generalization ability. In this study, a self-attention mechanism called GD-Attention is proposed, which considers global pixel value distribution information and guide the deep learning model to give more concern on the leaf disease area. Based on data augmentation, the proposed method inputs both the image and its pixel value distribution information to the model. The GD-Attention mechanism guides the model to extract features related to pixel value distribution information, thereby increasing attention towards the disease area. The model is trained and tested on the Plant Village (PV) dataset, and by analyzing the generated attention heatmaps, it is observed that the disease area obtains greater weight. The results achieve an accuracy of 99.97% and 27 MB parameters only. Compared to classical and state-of-the-art models, our model showcases competitive performance. As a next step, we are committed to further research and application, aiming to address real-world, complex scenarios.

Tomatoes are an important economic crop in China, and crop diseases often lead to a decline in their yield. Deep learning-based visual recognition methods have become an approach for disease identification; however, challenges remain due to complex background interference in the field and the diversity of disease manifestations. To address these issues, this paper proposes the SXA-YOLO (an improvement based on YOLO, where S stands for the SAAPAN architecture, X represents the XIoU loss function, and A denotes the AsDDet module) symmetric perception recognition model. First, a comprehensive symmetry architecture system is established. The backbone network creates a hierarchical feature foundation through C3k2 (Cross-stage Partial Concatenated Bottleneck Convolution with Dual-kernel Design) and SPPF (the Fast Pyramid Pooling module) modules; the neck employs a SAAPAN (Symmetry-Aware Adaptive Path Aggregation Architecture) bidirectional feature pyramid architecture, utilizing multiple modules to achieve equal fusion of multi-scale features; and the detection head is based on the AsDDet (Adaptive Symmetry-aware Decoupled Detection Head) module for functional decoupling, combining dynamic label assignment and the XIoU (Extended Intersection over Union) loss function to collaboratively optimize classification, regression, and confidence prediction. Ultimately, a complete recognition framework is formed through triple symmetric optimization of “feature hierarchy, fusion path, and task functionality.” Experimental results indicate that this method effectively enhances the model’s recognition performance, achieving a P (Precision) value of 0.992 and an mAP50 (mean Average Precision at 50% IoU threshold) of 0.993. Furthermore, for ten categories of diseases, the SXA-YOLO symmetric perception recognition model outperforms other comparative models in both p value and mAP50. The improved algorithm enhances the recognition of foliar diseases in tomatoes, achieving a high level of accuracy.