Improving Harvesting Efficiency and Sustainability Through Lightweight Visual Perception: COS-DETR for Resource-Constrained Tea-Harvesting Robots

Visual perception is a key challenge for tea-harvesting robots because tender buds must be distinguished from dense foliage under variable illumination and localized rapidly enough to support grasping on embedded hardware. We developed COS-DETR, a detector based on RT-DETR that incorporates a Faster CGLU block, the OmniKernel module with an internal Frequency–Spatial Attention Module (FSAM), and SPDConv. We also evaluated a compression pipeline that combines layer adaptive magnitude pruning (LAMP), channel pruning, and Mimic + Linear knowledge distillation. The uncompressed model achieved 87.1% mAP0.5, the highest value among the evaluated detectors. After pruning and distillation, the model achieved 85.7% mAP0.5 and 58.1 FPS on an NVIDIA Jetson Orin NX, with a 37.3% reduction in computational cost at the expense of only 1.4 percentage points of mAP0.5. In 15 static field trials, the integration of this detector with a stereo camera and a Delta manipulator yielded absolute positioning errors below 5 mm. As the harvesting platform remains an experimental prototype, these trials validated static target positioning only; continuous dynamic picking was not performed, so the picking success rate and the missed-detection rate under dynamic conditions are not reported in this study. These results indicate that a compressed vision transformer detector can provide the speed, accuracy, and positional precision required for robotic process control in precision tea harvesting within the static positioning scope evaluated here.