3D LULC classification using multispectral LiDAR and deep learning: current and prospective schemes

2026-05-21

AI Summary

This study addresses the lack of publicly available multispectral LiDAR datasets aligned with national mapping and cadastral agency (NMCA) classification standards, which has hindered research in 3D land use and land cover (LULC) classification. To bridge this gap, the authors introduce Loosdorf-MSL, the first dual-wavelength (532 nm / 1064 nm) multispectral LiDAR benchmark dataset tailored to NMCA requirements, along with a two-level LULC taxonomy (Level 1: 8 classes; Level 2: 20 classes). Comprehensive evaluations using seven state-of-the-art models, including Point Transformer V3, demonstrate that incorporating multispectral information improves mean Intersection over Union (mIoU) by 1.1 and 7.8 percentage points on Levels 1 and 2, respectively, achieving final performance of 79.4% and 58.9%. These results underscore the critical role of multispectral features in fine-grained material discrimination.

Abstract

Land Use Land Cover (LULC) classification is essential for national 3D mapping, geospatial analysis, and sustainable planning. Multispectral (MS) LiDAR provides synchronized spatial-spectral information, and deep learning (DL) enables 3D point cloud semantic segmentation; however, adoption is limited by the lack of publicly available urban and suburban MS LiDAR datasets aligned with National Mapping and Cadastral Agencies (NMCAs) classification schemes. This study addresses these gaps by introducing L1 and L2 NMCA-aligned LULC classification schemes and a new benchmark MS LiDAR dataset. We evaluate seven state-of-the-art DL models and perform spectral ablation studies at both levels of detail. Results show that Point Transformer V3 achieves the best performance, with mIoU of 79.4% (L1, 8 classes) and 58.9% (L2, 20 classes) using a dual-wavelength LiDAR system (532 nm and 1064 nm). Ablation results show that multispectral information improves performance over geometry-only inputs, with gains of 1.1 percentage points at L1 and 7.8 points at L2. These results highlight the value of LiDAR reflectance for fine-grained material discrimination and support the evolution of NMCA LULC schemes toward higher semantic detail. The Loosdorf-MSL dataset contributes a new benchmark for consistent national and international LULC mapping.

Problem

• 3D LULC classification
• multispectral LiDAR
• deep learning
• benchmark dataset
• NMCA-aligned schemes

Innovation

• multispectral LiDAR
• 3D LULC classification
• deep learning
• Point Transformer V3
• spectral ablation