PCL_ACC

This is an implemention using CUDA to accelerate the PCL algorithm including transform and + and some other.

PCL is a library for 2D/3D processing and analysis of point clouds,but it is slow in some algorithms.although I build PCL on CUDA, it is still slow than use CUDA to rewrite the algorithm. In this project, I rewrite part of PCL algorithm using CUDA and truely to make it faster.OUR code is not a complete rewrite of PCL, but it is a good example of how to use CUDA to accelerate PCL.

ChangeLog
Install
Usage
Contributing
License

ChangeLog

2024-02-27

completed PCL crop algorithm.

Attention:every variable in kernel function must copy to device.

PCL crop

2024-02-22

Establish the project.

PCL +
PCL transform
establish README.md

Install

install by download the source code and compile it.

requirements

This module depends upon the following libraries.

PCL
EIGEN
CUDA

Usage

TEST CODE

1、edit the CMakeLists.txt to add the path of PCL and CUDA
2、edit the PointCloud type in function.
3、mkdir build && cd build  then run the cmake and make.
4、./YourExecutable

Note: You should test the code first，and then you can use it as a library.

USE AS A LIBRARY

1、 if you uncomment the line `cuda_add_library(cuda_pcl_lib SHARED ${SOURCE_FILES})`,you should get a lib file `libcuda_pcl_lib.so` after the make.
2、 add the path of the lib file to YOUR PROJECT CMAKELISTS.txt
3、 copy cuda_pcl.h to YOUR PROJECT.
4、 INCLUDING the header file in your code,and use the cuda_pcl function. `#include "cuda_pcl.h"`
5、(optional) I would like to copy the libcuda_pcl_lib.so to `CUDA LIB PATH` and just include the CUDA PATH in cmakefiles,so dont need include the lib file path.

Contributing

See the contributing file!

PRs accepted.

Small note: If editing the Readme, please conform to the standard-readme specification.

mx2013713828/pcl_acc