GM-APD LiDAR target detection method based on trigger probability interval difference

Area of Science:

• Optoelectronics
• Signal Processing
• LiDAR Technology

Background:

• Geiger-mode avalanche photodiode (GM-APD) LiDAR systems face challenges with weak echo detection under strong background light.
• Existing methods struggle to maintain performance and accuracy in such noisy environments.

Purpose of the Study:

• To develop an advanced detection method for GM-APD LiDAR systems to overcome weak echo signal loss in high background light conditions.
• To improve the reliability and precision of LiDAR measurements.

Main Methods:

• A novel detection method utilizing trigger probability gradient statistics is proposed.
• Poisson statistics and an interval difference operator are employed to convert non-uniform noise into a stationary process.
• A 30% overlapping stepping gating strategy is implemented to boost transient responses and mitigate dead-time effects.

Main Results:

• The method achieves a 93.5% detection probability at a signal-to-background ratio of 0.1.
• A false-alarm rate below 1% was maintained under challenging conditions.
• Field tests demonstrated localization errors within 0.6 meters at a distance of 1060 meters.

Conclusions:

• The proposed trigger probability gradient statistics method significantly enhances weak echo detection in GM-APD LiDAR under strong background light.
• The technique offers robust performance, high detection probability, and sub-meter localization precision, validating its practical applicability.