High-Resolution Terrain Modeling Using Airborne LiDAR Data with Transfer Learning

Huxiong Li,Sarah Narges Fatholahi,Weikai Tan,Saied Pirasteh,Jun Liu,Jonathan Li,Weiya Ye

doi:10.3390/rs13173448

Huxiong Li, Sarah Narges Fatholahi + Show 5 more

Open Access

PDF Available

https://doi.org/10.3390/rs13173448

Copy DOI

Export

Save

Cite

Abstract
Highlights/Summary
Full-Text PDF
Similar Papers

Abstract

Listen

This study presents a novel workflow for automated Digital Terrain Model (DTM) extraction from Airborne LiDAR point clouds based on a convolutional neural network (CNN), considering a transfer learning approach. The workflow consists of three parts: feature image generation, transfer learning using ResNet, and interpolation. First, each point is transformed into a featured image based on its elevation differences with neighboring points. Then, the feature images are classified into ground and non-ground using ImageNet pretrained ResNet models. The ground points are extracted by remapping each feature image to its corresponding points. Last, the extracted ground points are interpolated to generate a continuous elevation surface. We compared the proposed workflow with two traditional filters, namely the Progressive Morphological Filter (PMF) and the Progressive Triangulated Irregular Network Densification (PTD). Our results show that the proposed workflow establishes an advantageous DTM extraction accuracy with yields of only 0.52%, 4.84%, and 2.43% for Type I, Type II, and the total error, respectively. In comparison, Type I, Type II, and the total error for PMF are 7.82%, 11.60%, and 9.48% and for PTD 1.55%, 5.37%, and 3.22%, respectively. The root means square error (RMSE) for the 1 m resolution interpolated DTM is only 7.3 cm. Moreover, we conducted a qualitative analysis to investigate the reliability and limitations of the proposed workflow.

Highlights

High-quality digital terrain models (DTMs) or digital elevation models (DEMs) are vital to various applications, such as urban building reconstruction [1], carbon storage estimation [2], off-ground object detection [3], and land cover mapping [4]
This study presented a workflow for DTM generation using airborne LiDAR data based on convolutional neural network (CNN) and transfer learning for a small area covering the main campus of the
This study presented a workflow for DTM generation using airborne LiDAR data based on CNN and transfer learning for a small area covering the main campus of the University of Waterloo region

Summary

Introduction

High-quality digital terrain models (DTMs) or digital elevation models (DEMs) are vital to various applications, such as urban building reconstruction [1], carbon storage estimation [2], off-ground object detection [3], and land cover mapping [4]. Raw LiDAR point clouds include both ground and non-ground points. After data geo-referencing, outlier removal and interpolation, the entire point cloud can be transformed into a digital surface model (DSM), including the elevation of both ground and non-ground objects. Off-ground objects, such as trees and buildings, are removed during the creation of DTMs. In Ontario, Canada, the DEMs generated by digital photogrammetry under the Southwestern Ontario Orthophotography Project have low vertical accuracy of 50 cm [5]. In 2018, Natural Resources Canada released the High-Resolution Digital Elevation Model products, which are derived from LiDAR [6]

Objectives

Results

Conclusion