An Improved Clustering-Based Denoising Method for Single Photon Point Cloud Data

Abstract

The invention proposes an improved method for denoising single-photon point cloud data based on clustering. The invention converts the three-dimensional point cloud data of the photon counting laser altimeter into two-dimensional point cloud data, and the two-dimensional point cloud data obtains the two-dimensional point cloud data after rough denoising through the rough denoising method; The homogenized two-dimensional point cloud data is obtained by homogenizing the point cloud data; the two-dimensional point cloud data is obtained by the fine denoising method through the homogenized two-dimensional point cloud data; Point cloud data is denoised and signal extracted at a faster speed and with higher recall and precision.

Description

technical field [0001] The invention belongs to the signal processing field of single-photon detection, and in particular relates to an improved cluster-based single-photon point cloud data denoising method. Background technique [0002] Single-photon detection technology is a new type of detection technology developed in recent years. Taking the counting spaceborne laser altimeter that has been used as an example, it uses single-photon detection devices as receiving devices, and its sensitivity is higher than that of traditional laser altimeters. 2-3 orders of magnitude, it is easier to achieve direct 3D imaging with micropulse, high repetition rate, and multiple beams, which has great technical advantages and development prospects. The photon-counting spaceborne laser altimeter has high measurement frequency, large amount of data, and various types of surface detection. Therefore, it is necessary to study a fast, effective, and robust point cloud data denoising method for ...

AI Technical Summary

Problems solved by technology

Among them, the grid-based method is prone to misjudgment when the target has severe fluctuations or steep slopes; the method based on partial statistical information has high time complexity and long calculation time, and due to the changeable actual terrain, it is difficult to select the density threshold; The clustering method is greatly affected by the uneven distribution of point clouds, and it is difficult to correctly extract targets with large reflectivity and environmental differences at the same time with a fixed threshold

Images