In this repository, we have implemented several point cloud processing algorithms in C++ and python. For each project, you will have an overview from the readme file, where we introduce the idea of the algorithms and some illustrations of the results.
Fig.1 PCA and Surface Normal of the point cloud
Fig.2 Above: RGB image; Middle: Depth image; Down: Upsampled depth image
My kdtree ------------
Kdtree: build 283.742ms, knn 11.855ms, radius 0.505ms, brute 13.218ms
My octree --------------
Octree: build 11727.653ms, knn 0.981ms, radius 0.732ms, brute 17.305ms
Fig.3 Benchmark of clustering algorithm.
Left 1: Mykmeans; Left 2: MyGMM
Left 1: Mykmeans; Left 2: MyGMM
Fig.4 Ground segmentation(RANSAC + Least square optimization). Blue part is the segmented ground
Fig.5 DBSCAN clustering algorithm of the point clouds after aplying ground filter
Fig.6 left: loss diagramm right: accuracy diagramm
Fig.7 P-R Curve of car detection and orientation
Fig.8 Feature detection of gituar and airplane
React
Typescript
Django
Laravel
Facebook
Microsoft
Google
Alibaba
D3
Tencent