逆向工程中复杂实物曲面重建的仿真研究

摘要

对逆向工程中复杂实物曲面进行重建,在分析复杂实物曲面点云方面具有重要意义.对复杂实物曲面重建时,由于寻找实物曲面点云的匹配位置难度大、精准度低,导致重建效果不理想.传统方法利用双B样条函数来确定点云最优匹配位置,但由于迭代次数过多,导致重建过程过于复杂.提出基于遗传理论的逆向工程中复杂实物曲面重建方法.结合SUSAN角点检测理论提取复杂实物曲面角点特征信息,对提取的角点特征进行曲面点云初始匹配得到初始匹配点集,以获得实物曲面点云最优匹配点集为依据,对待重建的曲面密集点云和参考点云进行匹配,利用实物曲面点云匹配的近似性测度作为遗传个体的适应度函数,得到实物曲面点云最优匹配位置,实现逆向工程中复杂实物曲面重建.实验结果表明,所提方法能够有效提升逆向工程中复杂实物曲面重建精度,且鲁棒性较优.%Traditional method uses double B-spline function to confirm optimal matching position of point cloud.However,iteration times are overmuch,which leads to complex reconstruction process.The authors of this article put forward a method for surface reconstruction of complex object in reverse engineering based on the genetic theory.Feature information of surface angular point integrated was extracted based on SUSAN angular point detecting theory.Initial matching of surface point cloud was carried out to the extracted angular point feature in order to obtain initial matching point set.The authors also matched dense point cloud of surface waiting for reconstruct and reference point cloud according to the optimal matching point set;then used similarity measure matching with surface point cloud as fitness function of genetic individual to obtain optimal matching position of surface point cloud.Thus,the surface reconstruction was achieved.Experimental results indicate that the method can improve the surface reconstruction precision effectively,and its robustness is excellent.