首页> 外文期刊>Information Technology Journal >A New Registration Method for Scattered Point Clouds from Multi-views
【24h】

A New Registration Method for Scattered Point Clouds from Multi-views

机译:多视角的散乱点云配准新方法

获取原文
           

摘要

Aiming at the multi-views scanning data of the object with complex curved surface, it is put forward that a new effective registration method from 3D armadillo range images to complete 3D model. Firstly, the coarse registration is made by using differential invariant features and ZNCC (zero-mean normalized cross-correlation coefficient), then ICS (iterative closest surface) is used to achieve accurate registration fast, final the complete 3D armadillo integrated model is got. By comparatively analyzing the registration result, registration error, number of iterations and convergence speed with numerical experiments, it is explained the validity and superiority of this method.
机译:针对复杂曲面物体的多视角扫描数据,提出了一种从3D犰狳距离图像到完整3D模型的有效配准方法。首先,利用微分不变特征和零均值归一化相关系数ZNCC进行粗配准,然后利用ICS(迭代最近表面)快速实现配准,最终得到完整的3D犰狳集成模型。通过数值实验比较分析了配准结果,配准误差,迭代次数和收敛速度,说明了该方法的有效性和优越性。

著录项

相似文献

  • 外文文献
  • 中文文献
  • 专利
获取原文

客服邮箱:kefu@zhangqiaokeyan.com

京公网安备:11010802029741号 ICP备案号:京ICP备15016152号-6 六维联合信息科技 (北京) 有限公司©版权所有
  • 客服微信

  • 服务号