Iterative TIN-based automatic filtering of sparse LiDAR data

Abstract

In this letter, a novel method for automatic separation of terrain points and object points using sparse Light Detection And Ranging (LiDAR) data is developed. The proposed method is based on iterative elimination of step edges connecting terrain points to object points. The first stage is to detect these edges. Using a triangulated irregular network (TIN) interpolation of the LiDAR raw points, each triangle is assigned to one of two classes: edge triangle or non-edge triangle, using the slope as the discriminative function. Edge triangles are located at the boundary between terrain and non-terrain points; therefore, the vertices of each triangle consist of terrain points and object points. Initially the lower points are considered as terrain points and the higher points as object points. The elevation of object points is adjusted using an interpolation method based on the estimated local slope. The local slope is calculated using non-edge adjacent triangles to the step triangle. The slopes of modified triangles are recalculated using the new elevation. This process is repeated until no triangle is assigned to the edge triangle class. At the end of this process, all the adjusted points are classified as object points and the other points as terrain points. Validation is done by computing the type I (terrain points misclassified as object points) and type II (object points misclassified as terrain points) errors. We used two large datasets containing many complex objects. The proposed approach achieved an overall accuracy higher than 90% and an average error of less than 10%.