Optimization of linear signal processing in photon counting lidar using Poisson thinning.

Area of Science:

• Geospatial science
• Optical remote sensing
• Signal processing

Background:

• Photon counting lidar (PCL) signals often contain random noise requiring smoothing.
• Standard smoothing techniques can introduce systematic errors by smearing high-gradient signals.
• Quantifying and balancing random versus systematic errors in PCL data is challenging and scene-dependent.

Purpose of the Study:

• To introduce a novel method for optimizing smoothing filter parameters in photon counting lidar.
• To enable quantitative evaluation and selection of filter parameters based on scene characteristics.
• To improve the accuracy of photon counting lidar signal processing.

Main Methods:

• Implementation of Poisson thinning for lidar signal processing.
• Quantitative evaluation of filter parameters using defined criteria.
• Optimization of smoothing based on scene-dependent error balancing.

Main Results:

• Poisson thinning allows for optimal selection of filter parameters for photon counting lidar.
• This method effectively balances random noise suppression with the minimization of systematic errors from signal smearing.
• The optimization process is computationally inexpensive and easy to implement.

Conclusions:

• Poisson thinning offers a robust solution for noise reduction in photon counting lidar.
• This approach enhances the reliability and accuracy of lidar data by managing systematic errors.
• The method provides a scene-adaptive strategy for processing photon counting lidar signals.