Hongduo Zhang

The University of Tokyo, Hunan Agricultural University

Papers

Total Citations

H-Index

About

Hongduo Zhang’s research lies at the intersection of agricultural robotics, computer vision, and precision automation, with a focus on developing intelligent picking systems for high-value crops. His work addresses the critical challenge of labor shortages in agriculture by enabling robots to accurately detect, localize, and harvest fruits in complex field environments. Zhang’s most cited paper (31 citations) introduces a coupled YOLO/Mask R-CNN framework for strawberry recognition using 3D binocular cameras, tackling issues of occlusion and varying maturity. He further advanced fruit detection with an improved YOLOv3 algorithm for citrus in cluttered orchards (15 citations), achieving fast and robust recognition. Beyond perception, Zhang has contributed to the mechanical design of harvesting robots through systematic parameter analysis of citrus stalk cutting (8 citations), optimizing blade speed and cutting angles for clean, efficient detachment. His integrated approach—combining deep learning-based vision with physical interaction modeling—has practical implications for reducing crop damage and improving harvest throughput. Zhang’s work is widely cited by researchers in agricultural robotics and precision farming, and his methodologies serve as a foundation for developing next-generation autonomous fruit pickers.