海洋测绘2026，Vol.46Issue(3)：16-19,24,5.DOI:10.3969/j.issn.1671-3044.2026.03.004
改进渐进三角网加密滤波算法在红树林的应用
Application of improved progressive triangulated network densification filtering algorithm in mangrove forests
摘要
Abstract
For mangrove-covered areas,LIDAR is difficult to penetrate,and there are water bodies in the area,which leads to the problem of seed points being easily misjudged by using traditional PTD filtering algorithms.This paper proposes a surface fitting method for seed points within the area to calculate the residuals of seed points,and uses whether the residuals of seed points satisfy the confidence interval to determine point attributions.In addition,a morphological filtering algorithm is introduced to eliminate water bodies and divide water boundaries through opening operations.The filtering classification results of the improved PTD filtering algorithm in this paper is compared with those of the traditional PTD filtering algorithm.Point clouds in the mangrove area are selected as data samples,and qualitative and quantitative analysis of the data results are conducted.The results show that compared to the traditional PTD algorithm,the improved algorithm reduces type I error rate from 3.0%to 2.6%,type II error rate from 8.3%to 6.1%,and total error rate from 6.3%to 4.8%.This provides a new research method and idea for subsequent terrain acquisition in mangrove areas.关键词
点云滤波/不规则渐进三角网加密/形态学滤波/置信区间/t分布Key words
point cloud filtering/progressive TIN densification/morphological filtering/confidence interval/t-distribution分类
天文与地球科学引用本文复制引用
王伟..改进渐进三角网加密滤波算法在红树林的应用[J].海洋测绘,2026,46(3):16-19,24,5.基金项目
天津水运工程研究院有限公司发展基金(SJY202506). （SJY202506）
