基于极小样本轻量化网络的物品检测抓取分拣平台研究

Research on Object Detection,Grabbing and Sorting Platform Based on Minimal Sample Lightweight Network

本文对物体抓取分拣相关平台进行研究,发现智能化的检测算法分拣装置是物流行业中的薄弱点,针对六自由度工业机器人,基于生成对抗网络构建轻量化神经网络模型,添加注意力机制,按照拆垛、检测、分拣、再码垛的流程设计具体技术路线图,并验证识别以及抓取的可行性。该设计在现代物流、智能装备等领域有广阔应用前景,有一定的研究价值。

This paper discusses the relevant platforms designed for object detection,grabbing and sorting.Through domestic research and analysis,it is found that the intelligent detection algorithm sorting device is the weak point in the logistics industry.Aiming at the 6-DOF industrial robot,a lightweight neural network model is built based on the generated confrontation network,and attention mechanism is added.The specific technical roadmap is designed according to the process of unpacking,detection,sorting and re-stacking,and the feasibility of identification and grabbing is verified.The design has broad application prospects in modern logistics,intelligent equipment and other fields,and has certain research value.