Scholarlink

Abstract <p>The high time consumption and storage requirements are the main obstacles for the current applications of point cloud data, especially for high-density data. To alleviate the computational burden while preserving the important geometric features, an effective point cloud simplification method based on the Harris-SIFT feature and improved voxel filtering is proposed, which divides the point cloud points into strong feature points and weak feature points. Firstly, the key points are extracted by the improved Harris algorithm from the original point cloud data, and the SIFT algorithm is used to divide them into the strong feature points and weak feature points. Then, the weak feature points are further simplified via the improved voxel filtering by introducing the scale factor, and finally, the strong and weak feature points are integrated to be the new data according to the structure of the point cloud. Experimental results of three well-known methods under four models of point cloud demonstrate that not only the important geometric features will be well preserved, but also the higher simplification rate will be obtained in terms of the information entropy and average information entropy.</p>

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

An Effective Point Cloud Simplification Method Based on the Harris-SIFT Feature and Improved Voxel Filtering

  • 期刊论文

  • 发表日期：

    2025年2月1日

查看全文
  • Chunxiang Liu,
  • Tianqi Cheng,
  • Xinping Guo,
  • YuWei Wang,
  • Lei Wang

摘要

Abstract

The high time consumption and storage requirements are the main obstacles for the current applications of point cloud data, especially for high-density data. To alleviate the computational burden while preserving the important geometric features, an effective point cloud simplification method based on the Harris-SIFT feature and improved voxel filtering is proposed, which divides the point cloud points into strong feature points and weak feature points. Firstly, the key points are extracted by the improved Harris algorithm from the original point cloud data, and the SIFT algorithm is used to divide them into the strong feature points and weak feature points. Then, the weak feature points are further simplified via the improved voxel filtering by introducing the scale factor, and finally, the strong and weak feature points are integrated to be the new data according to the structure of the point cloud. Experimental results of three well-known methods under four models of point cloud demonstrate that not only the important geometric features will be well preserved, but also the higher simplification rate will be obtained in terms of the information entropy and average information entropy.