基于遗传算法的三维模型匹配方法  

A 3D Models Matching Algorithm Based on Genetic Algorithm

作　　者：叶建华[1,2] 高诚辉[1] 江吉彬[2] 

机构地区：[1]福州大学机械工程及自动化学院,福建福州350108 [2]福建工程学院机械与汽车工程学院,福建福州350108

出　　处：《图学学报》2015年第1期22-27,共6页Journal of Graphics

基　　金：国家自然科学基金资助项目(51305079);福建省自然科学基金资助项目(2013J01168);福建省教育厅A类科技资助项目(JA13216);福建省省属高校科研资助项目(JK2012031)

摘　　要：逆向测量模型与正向设计模型的自动匹配是三维检测的关键技术之一。通过空间六自由度的旋转与平移变换调整模型方位,基于K-D树和拓扑信息获取三维模型与不同方位平面的相交轮廓。利用二维相交轮廓的差异度作为两模型间的匹配判据,避免海量数据点与复杂曲面间的直接匹配计算。采用遗传算法进行两模型最佳匹配方位的求解,以空间六自由度为个体的染色体,通过群体的多点搜索,历经选择、交叉、变异操作,得到全局最佳匹配方位。通过实例验证了方法的有效性。Marching mesh model which get by measurement sensor to CAD model is one of key technology in geometrical parameter measurement. It is necessary to research the consistent match orientation and position for two models. In this paper, a 3D models matching algorithm was proposed based on genetic algorithm. The translation and rotation operations in three dimensional spaces to adjust the position and orientation are used, and 2D outline features was used to match the two models. Genetic algorithm is applied to searching the consistent match orientation and position for two models. The prototype system was developed and the results were reported. Based on practice, this method is robust and efficient.

关 键 词：模型匹配 三维模型 遗传算法 归一化 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]