An optimized YOLO?based object detection model for crop harvesting system

The adoption of automated crop harvesting system based on machine vision may improve productivity and optimize the operational cost. scope this study is to obtain visual information at plantation which crucial in developing an intelligent system. This paper aims develop automatic detection with high accuracy performance, low computational cost lightweight model. Considering advantages YOLOv3 tiny, optimized tiny network namely YOLO-P proposed detect localize three objects palm oil include fresh fruit bunch, grabber tree under various environment conditions. model incorporated backbone densely connected neural network, multi-scale architecture anchor box size. experimental results demonstrated that achieved good mean average precision F1 score 98.68% 0.97 respectively. Besides, performed faster training process generated 76 MB. was also tested identify bunch maturities 98.91%. comprehensive show can effectively perform robust accurate plantation.