Method to Solve Underwater Laser Weak Waves and Superimposed Waves.

Area of Science:

• Geospatial technology
• Optical remote sensing
• Signal processing

Background:

• Lidar technology enables underwater terrain detection, but signal resolution remains challenging.
• Underwater laser echo signals are prone to weak and superimposed waves, complicating analysis.
• Existing waveform decomposition methods are ineffective for complex underwater signals, leading to data errors.

Purpose of the Study:

• To develop an improved waveform decomposition method for underwater Lidar signals.
• To enhance the accuracy and stability of underwater terrain detection.
• To address limitations in processing weak and superimposed underwater laser echo signals.

Main Methods:

• Utilized a drone with a 532 nm laser for pond terrain detection.
• Proposed an improved inflection point selection decomposition method for parameter estimation.
• Applied oscillating particle swarm optimization (OPSO) and Levenberg-Marquardt (LM) algorithm for parameter optimization.

Main Results:

• The proposed method demonstrated superior performance with RMSE of 2.544 and R² of 0.995975 compared to other methods.
• Optimized parameters resulted in waveforms closely matching the original signal.
• Decomposition of simulated complex waveforms yielded an RMSE of 0.0016 and R² of 1, with Gaussian components fully representing the original waveform.

Conclusions:

• The improved inflection point selection decomposition method is highly accurate and stable for underwater Lidar waveform analysis.
• This method effectively processes complex waveforms, including weak and superimposed signals, outperforming existing techniques.
• The study provides a robust solution for accurate underwater terrain detection using Lidar technology.