控制理论与应用2013，Vol.30Issue(6)：724-733,10.DOI:10.7641/CTA.2013.20941
基于三维点云的阶梯目标检测与参数估计
Stair detection and parameter estimation using 3D point cloud
摘要
Abstract
Stair detection and 3D modeling are of great significance for autonomous navigation and motion planning of mobile robots.Considering the structure diversity of stairs and the distribution uncertainty of point cloud in practical application,we proposed an adaptive stair detection and parameter estimation method based on stair topology model and fuzzy set theory.By employing the topology of stair profile and a histogram algorithm,we improve the accuracy and robustness of the stair edge location estimation.Adopting the strategy of in-level extraction and cross-level linking,we construct a set of candidate edge lines.On the basis of this result,the cascade probability of candidate edge lines is estimated by using fuzzy transformation and adaptive fuzzy inference.A simulated annealing algorithm is applied to the global optimum search for the candidate edge-line combination to validate the estimation of 3D stair modeling parameters.Experimental results and data analysis show the validity and feasibility of the proposed approaches.关键词
三维点云/阶梯目标检测/模糊推理/模拟退火Key words
3D point cloud/ stair detection/ fuzzy inference/ simulated annealing分类
信息技术与安全科学引用本文复制引用
钟晨,王伟,庄严..基于三维点云的阶梯目标检测与参数估计[J].控制理论与应用,2013,30(6):724-733,10.基金项目
国家自然科学基金资助项目(61075094,61035005) （61075094,61035005）
机器人学国家重点实验室基金资助项目(RLO200911). （RLO200911）
