Support
What is the recommended workflow for generating a Digital Terrain Model (DTM) from LiDAR data?
The transformation of raw LiDAR data into an accurate digital model is a fundamental process in geospatial engineering that requires a sequence of technical phases to ensure terrain fidelity. The optimized workflow in Tcp PointCloud Editor is divided into four main stages:
Point Cloud Cleaning Original data often contains anomalies due to sensor limitations or environmental conditions. It is essential to apply debugging tools to obtain a dense and clean dataset:
Selection by Isolation: this command allows you to identify and delete noisy points or small groups of isolated points that do not belong to the actual terrain surface.
Selection by Return: used to filter points based on the order of the received echo; usually, the first returns (which typically correspond to vegetation) are discarded to keep the last returns that better define the ground.
Ground Point Classification After cleaning, the objective is to isolate points that strictly belong to the natural surface. To do this, the software offers two main approaches:
Selection by CSF (Cloth Simulation Filter): a highly efficient algorithm that simulates a flexible cloth falling over the inverted cloud to automatically extract the ground relief.
AI Classification: uses Artificial Intelligence models trained for outdoor scenes to automatically label terrain versus other elements such as buildings, vehicles, or tall vegetation.
DTM Generation Once the "Ground" category is isolated, proceed with geometric modeling:
Create Surface: generates an irregular triangle network (TIN) from points classified as ground.
Create Grid: converts the model into a regular grid (Digital Elevation Map) with an ideal cell size, also allowing holes to be filled by interpolation.
Validation and Analysis Finally, it is essential to evaluate the quality of the model using inspection tools:
Create Contours: allows you to generate contour lines to detect visual anomalies or peaks in the relief.
Visual Analysis: application of configurable symbologies by elevations, slopes, or aspects to perform an advanced geographic study of the terrain directly in the viewer.
More information: How to generate a Digital Terrain Model from a LiDAR point cloud (news on aplitop.com).
Related videos
