项目名称: 巡线机器人运动模型辅助立体视觉导航
项目编号: No.51307003
项目类型: 青年科学基金项目
立项/批准年度: 2014
项目学科: 电工技术
项目作者: 李丹
作者单位: 安徽工业大学
项目金额: 25万元
中文摘要: 应用立体视觉技术在复杂自然背景下精确提取输电线路和其主要部件,并实现巡线机器人自主导航和越障定位,这在国内外都是一个新颖的前沿课题。本项目研究内容主要包括:双目摄像机快速自标定算法、复杂环境下的三维图像增强和去噪算法、输电线路主要部件特征取及自动识别、基于运动模型辅助的高效立体视觉导航技术、基于自适应信息融合的输电线越障定位技术。本项目着重探索巡线机器人在视觉导航和越障策略等研究方面的新思路和新方法,研究结果将对立体视觉导航技术的推广应用,对输电线路巡检技术的发展起到一定的促进作用。有很好的科研价值与工程应用前景。
中文关键词: 巡线机器人;视觉导航;运动模型;立体匹配;三维重建
英文摘要: The application of stereo vision technology in the context of accurate extraction of the transmission line and its major components from complicated nature background is a new frontier topics, this technique can be used for the inspection robot to travel along the wires and navigate obstacles. This project mainly include: the binocular camera fast self-calibration algorithm, 3D image enhancement and de-noising algorithm under the complex environment, the main components of the transmission line feature extraction and automatic identification, efficient stereo vision navigation technology based on motion model assisted, transmission lines obstacle navigated and position based on adaptive information fusion technology. The project focused on exploring inspection robot for new ideas and new ways in visual navigation and obstacle crossing strategies, the results will promote the application of three-dimensional visual navigation technology, the transmission line inspection technologies. It has good scientific significance and application prospects.
英文关键词: Inspection Robot;Vision Navigation;Motion Model;Stereo Matching;3D Reconstruction