A method and system for denoising imaging of lidar data by geiger apd array
Patent Information
- Application Number
- CN202510412253.8
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2025-04-03
- Publication Date
- 2026-10-16
AI Technical Summary
然而基于点云滤波的方法和基于波形识别的方法想要达到较好的去噪成像效果，往往涉及大量特殊算子的计算，计算复杂度较高
[0030] 1. This invention provides a denoising imaging method for Geiger APD array lidar data. Considering that the noise distribution of Geiger APD array lidar data generally has high temporal stability, this invention performs noise distribution estimation every preset time interval. The noise distribution of the lidar data is estimated in advance by curve fitting the frequency distribution of the global time count values of the lidar data sequence S to be denoised or the lidar data sequence S' collected within a preset time interval before the current lidar data to be denoised. Then, denoising imaging is performed on the lidar data sequence S based on the pre-estimated noise distribution, resulting in good denoising imaging performance. This invention does not involve the calculation of special operators, has low computational complexity, and only requires one noise distribution estimation within a certain time interval to meet the denoising requirements for that time interval, greatly reducing computational overhead. At the same time, this invention does not require training with large-scale and diverse training datasets, resulting in high real-time performance. Based on this, this invention can achieve good denoising imaging performance with low computational complexity while meeting real-time requirements.