LiDAR Echo Gaussian Decomposition Algorithm for FPGA Implementation

Area of Science:

• Geospatial technology
• Signal processing
• Computer engineering

Background:

• Existing LiDAR echo decomposition algorithms are computationally intensive and slow.
• Efficient processing of LiDAR full-waveform data is crucial for various applications.

Purpose of the Study:

• To develop and evaluate an FPGA-based improved Gaussian full-waveform decomposition method.
• To enhance the speed and efficiency of LiDAR echo decomposition without sacrificing accuracy.

Main Methods:

• Proposed an FPGA architecture with three modules: pre-processing, inflection point calculation, and Gaussian component parameter solution.
• Implemented pipelined data reading, Gaussian filtering, and second-order differential operations.
• Validated the method using LiDAR datasets from the Congo and Antarctic regions.

Main Results:

• The FPGA-based method demonstrates accuracy equivalent to PC-based processing.
• Achieved a processing speed enhancement of 292 times compared to PC-based methods.
• Successfully decomposed LiDAR full-waveform data with high efficiency.

Conclusions:

• The proposed FPGA-based Gaussian decomposition method offers a significant speedup for LiDAR data processing.
• This approach maintains high accuracy, making it a viable alternative for time-sensitive applications.
• FPGA acceleration is a promising direction for advancing LiDAR data analysis capabilities.