Point cloud bilateral filtering algorithm (with matlab code)

1. Overview of principles

In the field of two-dimensional images, the bilateral filtering algorithm corrects the position of the current sampling center point by considering the distance from the center pixel to the neighborhood pixel (one side) and the weight determined by the pixel brightness difference (the other side), thereby achieving Smooth filtering effect. At the same time, some adjacent sampling points that are too "difference" from the current sampling point will be selectively eliminated to achieve the purpose of maintaining the original features. In the three-dimensional point cloud, we replace the bilateral weights with the current point and neighboring points. The distance weight (one side) and the projection weight of the neighboring point along the normal vector of the current point (one side) are enough.

2. Implement the code

BilateralFiltering.m

%点云双边滤波
clear
close all;
clc
%获取点云数据
[fileName,pathName]=uigetfile('*.pcd','Input Data-File'