A revised progressive TIN densification for filtering airborne LiDAR data. (July 2017)

• Record Type: Journal Article
• Authors: Nie, Sheng; Wang, Cheng; Dong, Pinliang; Xi, Xiaohuan; Luo, Shezhou; Qin, Haiming
• Is Part Of: Measurement. Volume 104(2017)
• Page Start: 70
• Page End: 77
• Publication Date: 2017-07
• Subjects: LiDAR -- Filtering -- Triangular irregular network -- Progressive TIN densification -- Ground points; Weights and measures -- Periodicals; Measurement -- Periodicals; Measurement; Weights and measures; Periodicals; 530.8
• DOI: 10.1016/j.measurement.2017.03.007
• Languages: English
• ISSNs: 0263-2241

Abstract

Abstract: Filtering is an essential post-processing step for various applications of Light Detection and Ranging (LiDAR) data. Progressive triangular irregular network (TIN) densification (PTD) is a commonly used algorithm for filtering airborne discrete-return LiDAR data. However, this method has limitations in removing point clouds belonging to lower objects and preserving ground measurements in topographically complex areas. Therefore, this study revised the classic PTD method by building an improved TIN and changing the original iterative judgment criterions for better filtering airborne LiDAR point clouds. Similar to the classic PTD method, our revised PTD method also consists of three core steps: parameter specification, seed point selection and initial TIN construction, and iterative densification of TIN. To evaluate the performance of our revised PTD method, it was applied to benchmark datasets provided by ISPRS Working Group III/3, and compared with the classic PTD method in filtering airborne LiDAR data. Experimental results indicated that, our revised PTD approach performed better than the classic PTD method in preserving ground points in steep areas and removing non-ground points which belong to lower objects. Additionally, results showed that our revised PTD method is capable of reducing Type I errors, Type II errors and total errors by 10.26%, 0.79% and 8.07% respectively. Our revised PTD method offers a better solution for filtering airborne LiDAR discrete-return data.