CAJ | 학술논문

点云数据配准是三维激光数据处理的关键问题,结合曲率进行配准是点云配准非常有效的一种方式.为了提高配准的速度和精度,针对点云曲率和RANSAC算法的特点和应用,分析了两种方法之间的优缺点.研究了一种结合曲率的RANSAC点云配准算法,并利用改进的ICP算法对点云进行精确配准.结果表明:基于曲率的RANSAC算法能够显著提高初始配准的速度和精度,同时改进的ICP算法比原有ICP算法提高了二次配准的速度和精度.
점운수거배준시삼유격광수거처리적관건문제,결합곡솔진행배준시점운배준비상유효적일충방식.위료제고배준적속도화정도,침대점운곡솔화RANSAC산법적특점화응용,분석료량충방법지간적우결점.연구료일충결합곡솔적RANSAC점운배준산법,병이용개진적ICP산법대점운진행정학배준.결과표명:기우곡솔적RANSAC산법능구현저제고초시배준적속도화정도,동시개진적ICP산법비원유ICP산법제고료이차배준적속도화정도.
@@@@The point cloud data registration is the key issue of 3-d laser data processing .The point cloud registration combined with curvature distribution is a very effective away .In order to improve the registration speed and precision ,in view of the characters and application of point cloud curvature and RANSAC algorithm ,the advantages and disadvantages of the two methods are analyzed here .Then ,a RANSAC point cloud registration algorithm combined with curvature is studied ,and the improved ICP algorithm is used for point cloud registration .The results show that :the RANSAC algo-rithm based on curvature can significantly improve the initial registration speed and precision ,at the same time ,the im-proved ICP algorithm can significanlly improve the secondary registration speed and precision compared with the original ICP algorithm .