Scholarlink

Point cloud completionPoint cloud completion addresses the challenging task of reconstructing complete 3D geometries from partial scans. Existing methods normally ignore the effectiveness of viewpoint informationViewpoint information and adopt task-independent Farthest Point Sampling (FPS) to down-sample point clouds, leading to performance sacrifice.

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Object Completion from Point Clouds

  • 章节

  • 发表日期：

    2026年

查看全文
  • Yulan Guo,
  • Sheng Ao,
  • Zhiheng Fu,
  • Hao Liu,
  • Qingyong Hu

摘要

Point cloud completionPoint cloud completion addresses the challenging task of reconstructing complete 3D geometries from partial scans. Existing methods normally ignore the effectiveness of viewpoint informationViewpoint information and adopt task-independent Farthest Point Sampling (FPS) to down-sample point clouds, leading to performance sacrifice.