Powdery mildew has become a significant disease affecting the yield and quality of rubber trees in recent years. It typically manifests on the leaf surface at an early stage, rapidly infecting and spreading throughout the leaves. Therefore, early detection and intervention are essential to reduce the resulting losses due to this disease. However, the conventional methods of disease detection are both time-consuming and labor-intensive. In this study, we proposed a novel deep-learning-based approach for detecting powdery mildew in rubber trees, even in complex backgrounds. First, to address the lack of existing datasets on rubber tree powdery mildew, we constructed a dataset comprising 6200 images and 38,000 annotations. Second, based on the YOLO framework, we integrated a multi-scale fusion module that combines a Feature Focus and Diffusion Mechanism (FFDM) into the neck of the detection architecture. We designed an overall focus diffusion architecture and introduced a Dimension-Aware Selective Integration (DASI) module to enhance the detection of small powdery mildew targets, naming the model PM-YOLO. Furthermore, we proposed an automatic grading detection algorithm to evaluate the severity of powdery mildew on rubber tree leaves. The experimental results demonstrated that the proposed method achieved 86.9% mean average precision (mAP) and 85.6% recall, which outperformed the standard YOLOv10 by 7.6% mAP and 8.2% recall. This approach offered accurate and real-time detection of powdery mildew rubber trees, providing an effective solution for early diagnosis through automated grading.

Powdery mildew, caused by Erysiphe quercicola, is one of the primary diseases responsible for the reduction in natural rubber production in China. This disease is a typical airborne pathogen, characterized by its ability to spread via air currents and rapidly escalate into an epidemic under favorable environmental conditions. Accurate prediction and determination of the prevention and control period represent both a critical challenge and key focus area in managing rubber-tree powdery mildew. This study investigates the effects of spore concentration, environmental factors, and infection time on the progression of powdery mildew in rubber trees. By employing six distinct machine learning model construction methods, with the disease index of powdery mildew in rubber trees as the response variable and spore concentration, temperature, humidity, and infection time as predictive variables, a preliminary predictive model for the disease index of rubber-tree powdery mildew was developed. Results from indoor inoculation experiments indicate that spore concentration directly influences disease progression and severity. Higher spore concentrations lead to faster disease development and increased severity. The optimal relative humidity for powdery mildew development in rubber trees is 80% RH. At varying temperatures, the influence of humidity on the disease index differs across spore concentration, exhibiting distinct trends. Each model effectively simulates the progression of powdery mildew in rubber trees, with predicted values closely aligning with observed data. Among the models, the Kernel Ridge Regression (KRR) model demonstrates the highest accuracy, the R² values for the training set and test set were 0.978 and 0.964, respectively, while the RMSE values were 4.037 and 4.926, respectively. This research provides a robust technical foundation for reducing the labor intensity of traditional prediction methods and offers valuable insights for forecasting airborne forest diseases.