摘要
针对地形复杂、起伏较大的山区激光点云滤波精度低的难题,在深入理解不规则三角网(TIN)渐进加密算法基础上,首先对原始点云数据进行去噪和滤波处理,然后定量分析和比较不同参数阈值下的滤波效果。结果表明,参数阈值对基于TIN的滤波算法结果影响较大,需要结合具体的研究区域设置合适的阈值,普适性和自适应性强的滤波算法仍是点云滤波研究的重要内容。
Due to complex topography and undulating terrain, it is always a challenge to filter point clouds accurately. Based on the principle of progressive TIN densification algorithm, denoising and filtering were firstly applied to the original point clouds data. Then, we analyzed and compared the filtering effects under different parameter thresholds quantitatively. The experimental results show that the parameter threshold has great influence on the precision of filtering algorithm based on TIN and it is indispensable to set appropriate thresholds for specific research areas. In conclusion, it is necessary to study point cloud filtering algorithm based on TIN with strong universality and adaptability.
出处
《地理空间信息》
2019年第10期29-31,I0006,共4页
Geospatial Information
基金
国家自然科学基金资助项目(41671434)
基于光子计数激光雷达数据的森林生物量反演研究
河南省自然科学基金资助项目(182300410111)
河南省高等学校重点科研项目(18A420001)
关键词
TIN
点云数据
滤波
山区
定量评价
TIN
point cloud data
filtering
mountainous area
quantitative assessment
作者简介
第一作者:李奇翰,硕士研究生,主要从事摄影测量及遥感、激光雷达三维重建等方面的研究。
