Voxel-based spatial elongation filtering method for airborne single-photon LiDAR data.

Area of Science:

• Geospatial Science
• Remote Sensing Technology
• Signal Processing

Background:

• Airborne single-photon lidar (SPL) data is crucial for high-resolution 3D mapping.
• Noise in SPL data can degrade accuracy and hinder downstream applications.
• Existing filtering methods may not sufficiently address noise in dense SPL point clouds.

Purpose of the Study:

• To introduce a novel voxel-based spatial elongation filtering method for SPL data.
• To enhance signal-to-noise ratio in airborne lidar datasets.
• To develop a robust performance evaluation index for lidar noise filtering.

Main Methods:

• A novel voxel-based spatial elongation filtering technique was developed.
• Six adjacent points were generated for each data point to analyze local density.
• A predefined threshold within voxels was used to distinguish signals from noise.
• A filter performance evaluation index, incorporating false alarm and signal loss rates, was introduced.

Main Results:

• The proposed method demonstrated a reduction in noise within SPL data.
• The average false alarm rate achieved was 3.5%.
• This represents an 18.6% improvement over the traditional voxel-based spatial filtering method, which had an average false alarm rate of 4.3%.

Conclusions:

• The novel voxel-based spatial elongation filtering method is effective in reducing noise in airborne SPL data.
• The proposed method offers superior performance in minimizing false alarms compared to conventional voxel-based techniques.
• The developed evaluation index provides a comprehensive assessment of filter performance.