Research on Improving Laser Ranging Accuracy in Adverse Weather Conditions Using Artificial Intelligence Noise Reduction Technology

This study investigates artificial intelligence techniques to enhance laser ranging accuracy under adverse weather conditions. Conventional laser ranging systems suffer performance degradation in fog, rain, snow and dust-laden environments due to signal attenuation, scattering and backscatter effects. A deep learning noise-reduction system was developed, combining convolutional neural networks and adaptive filtering. The system identifies genuine laser return signals from background noise through a double-stage approach integrating signal preprocessing with a neural network trained on diverse meteorological conditions. Experimental verification was conducted in both laboratory simulations and real-field measurements. The results show the AI system achieves 78.3% improvement in ranging accuracy under dense fog and 63.7% under moderate rain compared to traditional techniques, maintaining sub-centimetre accuracy at short ranges and sub-decimetre accuracy at ranges up to 300 meters, with latency below 15 milliseconds. This work has implications for autonomous navigation, surveying and meteorological monitoring under variable environmental conditions.