Addressing the issue of fruit recognition and localization failures in harvesting robots due to severe occlusion by branches and leaves in complex orchard environments, this paper proposes an occlusion avoidance method that combines a lightweight YOLOv8n model, developed by Ultralytics in the United States, with active perception. Firstly, to meet the stringent real-time requirements of the active perception system, a lightweight YOLOv8n model was developed. This model reduces computational redundancy by incorporating the C2f-FasterBlock module and enhances key feature representation by integrating the SE attention mechanism, significantly improving inference speed while maintaining high detection accuracy. Secondly, an end-to-end active perception model based on ResNet50 and multi-modal fusion was designed. This model can intelligently predict the optimal movement direction for the robotic arm based on the current observation image, actively avoiding occlusions to obtain a more complete field of view. The model was trained using a matrix dataset constructed through the robot’s dynamic exploration in real-world scenarios, achieving a direct mapping from visual perception to motion planning. Experimental results demonstrate that the proposed lightweight YOLOv8n model achieves a mAP of 0.885 in apple detection tasks, a frame rate of 83 FPS, a parameter count reduced to 1,983,068, and a model weight file size reduced to 4.3 MB, significantly outperforming the baseline model. In active perception experiments, the proposed method effectively guided the robotic arm to quickly find observation positions with minimal occlusion, substantially improving the success rate of target recognition and the overall operational efficiency of the system. The current research outcomes provide preliminary technical validation and a feasible exploratory pathway for developing agricultural harvesting robot systems suitable for real-world complex environments. It should be noted that the validation of this study was primarily conducted in controlled environments. Subsequent work still requires large-scale testing in diverse real-world orchard scenarios, as well as further system optimization and performance evaluation in more realistic application settings, which include natural lighting variations, complex weather conditions, and actual occlusion patterns.

Accurate frost detection on leaf surfaces is critical for agricultural monitoring, yet existing methods struggle with segmentation errors caused by complex backgrounds (blurred, soil, weeds) and subtle frost-leaf texture differences. To address this, we propose MCGE-Frost, a multi-component gradient enhancement method that integrates color space analysis with gradient fusion theory. The algorithm extracts gradient features from individual color channels (HSV, Lab), applies adaptive weighting to enhance frost-leaf boundary contrast, and employs morphological filtering to suppress background noise. Experiments on leaf images demonstrate that MCGE-Frost achieves a total algorithmic error segmentation rate of 3.29%, significantly outperforming ExG (8.63%), OTSU (8.98%), and HSV (11.98%). The method reduces computational complexity by 40% compared to deep learning-based approaches while maintaining robustness across diverse backgrounds. MCGE-Frost achieves 0.8 s/image processing on GPU-accelerated systems, balancing accuracy and efficiency for edge deployment. Additionally, it improves the intelligence of frost quantification with minor manual calibration. This advancement supports real-time frost monitoring in precision agriculture, providing actionable insights for frost protection and crop management.

Accurate and in-situ monitoring of frost growth on plant leaves is crucial for disaster prevention in smart agriculture. To address the limitations of traditional methods in quantification and continuity, this study proposes a novel monitoring paradigm integrating dynamic dielectric spectrum analysis with hybrid intelligent algorithms. A mesh-electrode-based capacitive sensor was designed to capture in-situ and continuous dielectric spectrum changes on leaf surfaces. Subsequently, a hybrid SWT-SSA-LSTM model was constructed for high-fidelity denoising and prediction of the original signals. Field experiments demonstrated that this system could quantify frost layer mass and thickness with high precision. The established nonlinear regression models achieved coefficients of determination of 0.924 and 0.975, respectively. The prediction model exhibited outstanding performance, with a root mean square error as low as 1.475. This study establishes a complete technical closed-loop from physical perception to intelligent prediction, providing an innovative solution for precise frost monitoring in agriculture.

To address the poor mechanical adaptability of conventional equipment to 40 cm narrow-row sesame cultivation and the high weeding resistance and energy consumption of traditional weeding tools, this study developed an integrated bionic weeding and plant protection vehicle. The vehicle features a modular structure capable of three-row weeding and four-row plant protection, coupled with an extended-range hybrid powertrain. Its parallel linkage design enables terrain adaptation, ensuring consistent weeding depth of 3–6 cm and stable spraying height. Combined with an adjustable spraying width and a “detection–feedback–adjustment” mechanism to prevent plant collisions, the vehicle is fully compatible with the agronomic requirements of narrow-row cultivation. Inspired by mole cricket forelegs, the vehicle’s bionic weeding wheel blade model incorporates quantified biological features: quadratically fitted claw toe contours (R² > 0.97), a toe base height-to-width ratio of 1:2, and a toe groove radius-to-toe height ratio of 1:1. This design achieves a reliable biological-to-engineering translation. EDEM-based Discrete Element Method (DEM) simulations confirm that the bionic wheel outperforms conventional designs: the average torque is 17.4% lower (7.75 vs. 9.38 N·m), the soil disturbance rate is 8.2 percentage points higher (95.2% vs. 87.0%), and soil particle motion is more ordered (average velocity: 0.52 vs. 0.58 m/s), effectively reducing energy waste and improving weeding efficiency.