基于神经网络的智能型采棉机器人开发与研究  

Development and Study of Intelligent Cotton Picking Robot Based Upon Neural Networks

作　　者：袁新芳[1] 袁建宁[1] 于滨[1] 

机构地区：[1]南京工程学院机械工程学院,南京211167

出　　处：《农机化研究》2008年第2期88-91,共4页Journal of Agricultural Mechanization Research

基　　金：江苏省高校自然科学研究计划项目(05KJD210085);先进数控技术江苏省实验开放基金项目(KX105012;KXJ06123)

摘　　要：棉花成熟度和空间位置的识别是智能型采棉机器人研究的关键技术。为此,运用彩色图像处理、神经网络理论以及机器人技术研究开发了采棉机器人样机。该样机具有学习功能,采摘时根据示教的成熟棉花与未成熟棉花的标准样本自动摘取成熟棉花。同时,提出了基于神经网络的棉花成熟度判别策略及棉花重心位置的提取方法。For the intelligent cotton harvesting robot, the degree of maturation and recognition of the dimensional orientation are the key technologies. This investigation presented the development of a robotic prototype for picking cotton using color image processing, neural network, and robotic techniques. The developed prototype has learning ability and is able to pick the ripe cotton automatically according to the standard teaching patterns of both ripeand unripe cottons. A determining strategy of the cotton maturation based on neural network, and a ascertaining approach of the cotton barycenterare introduced.

关 键 词：农业工程 采棉机器人 设计 神经网络 图像处理 成熟度 

分 类 号：S225.911[农业科学—农业机械化工程] S126[农业科学—农业工程]