农用车辆视觉导航路径识别方法_赵博.pdfVIP

28 6 ( ) V o.l 2 8 N o. 6 2007 11 Journal of J iang su U n iversity( N a tura l Sc ience Ed ition) N ov. 2007 赵 博 朱忠祥 宋正河 谢 斌 毛恩荣 ( 100083 ) : 为提高农用车辆视觉导航系统对不同环境 适应性 研究了田间地头天空以及树木等复杂 环境对路径识别 影响. 提出用 一种新 中心线检测算法来取代原有 细化算法; 用腐蚀算法代替 原有 中值滤波; 据此设计了适合复杂环境 路径识别方法. 大量 田间试验证明 该方法对普通 环境和复杂环境均具有较好 适应性 能够快速可靠准确地提取导航路径特征. : 农用车辆; 路径识别; 视觉导航; 中心线检测; 腐蚀算法 + : TP242. 6 2 : A : 1671- 7775 ( 2007) 0 6- 0482- 05 Path recognition m ethod for vision navigation of agricultural vehicle ZHAO Bo, ZH U Zhong-xiang, SONG Zheng-he, XIE B in, MAO En-rong ( C ollege of Eng ineering Ch ina A gricultural U n ivers ity Be ijing 100083 C h ina) Abstract: In order to mi prove the adaptability o f the v ision nav igat ion system to various env ironm ents for the ag ricultura l veh icles the influ ence of the com p licated env ironm ent on the p ath recognition w as inves- t igated. A new algorithm of the center line de tect ing w as proposed instead o f the trad it iona l th inn ing algo- rithm; the erosion algorithm w as applied instead of the trad it iona l m edium f iltering algor ithm. A new m ethod for the path recogn ition under the com plicated env ironm ents w as designed. Large fie ld expermi en- tal resu lts prove th at the new m ethod is effective on bo th smi ple and com plicated env ironm ents and can obta in the target nav igat ion path rap idly re liab ly and accurate ly. K ey w ord s: agr icu ltura l veh icle; p ath recogn ition; vision nav