Zhen Wang

Underground pests are difficult to monitor because of their hidden activity, and traditional manual surveys are inefficient. Computer vision provides a promising approach for efficient detection; however, pest boundaries and textures in complex soil backgrounds can be obscured or confounded by soil particles, residual roots, shadows, and other factors, thereby reducing detection accuracy. In this study, an image dataset of grubs, cutworm larvae, and cutworm pupae was constructed, and a lightweight detection model, HFU-YOLO, was proposed using YOLO11n as the baseline. A Haar Wavelet Downsampling (HWD) module was introduced to reduce the loss of shallow edge and texture information. A Lightweight Frequency-Adaptive Dilated Convolution (LFADC) module was designed to enhance target feature representation and reduce soil background interference. A UGS-Lite auxiliary loss function was adopted to improve the localization stability of weak-boundary targets. The results showed that HFU-YOLO achieved Precision, Recall, mAP@0.5, and mAP@0.5:0.95 values of 77.6%, 78.3%, 83.5%, and 44.1% on the test set, respectively, representing improvements of 4.2, 3.2, 6.1, and 6.1 percentage points over YOLO11n. With 2.59 M parameters and 6.2 GFLOPs, HFU-YOLO provides a reference for intelligent recognition of underground pests in complex soil backgrounds.

Soil-borne diseases and continuous cropping obstacles have become critical factors constraining sustainable agricultural development. Traditional pesticide application methods commonly suffer from uneven dosage distribution, significant chemical waste, and environmental pollution. To enhance the precision and system stability of soil disinfection, this paper designs a precision pesticide application system for soil disinfection based on the Particle Swarm Optimization Proportional-Integral-Derivative algorithm (PSO-PID). Centered on a C37 controller, the system integrates the application pipeline, pumps, electric ball valves, multiple sensors, and a control terminal. The PSO-PID algorithm was deployed on the Codesys V2.3 platform, achieving precise flow control by adjusting electric ball valve openings in conjunction with velocity feedback. Simulink simulations showed that the PSO-PID algorithm outperforms conventional PID control in terms of settling time, overshoot, and steady-state error. Bench tests further validated the effectiveness of the proposed algorithm. Compared to conventional PID, the PSO-PID algorithm demonstrates superior control accuracy, faster response time, and enhanced application stability, with relative errors of 2.33%, 1.25%, and 1.20% respectively, while those of the conventional PID algorithm reach 3.67%, 3.35% and 4.88% respectively, representing a reduction of approximately 50% compared with the conventional PID algorithm. The results of the system application uniformity test indicated that the PSO-PID algorithm reduces relative error by approximately 40% compared to conventional PID, with the coefficients of variation being 2.02%, 1.73% and 1.81% respectively, which represented a significant improvement over those of the conventional PID algorithm (3.36%, 3.13% and 3.81%). Both application uniformity and stability outperform conventional PID algorithms, effectively minimizing application deviation. It outperforms conventional PID in both application uniformity and stability, effectively minimizing application deviation. The findings demonstrate that the proposed PSO-PID application control method achieves high control accuracy and application stability, providing reliable technical support for precision soil disinfection application.