💡Algorithms for Obstacle/Object detection using traditional point cloud processing methods.
The current code version has completed the basic algorithm implementation and can be used as a baseline The method used:
- Filter acceleration
- Clustering
- Detection
- Tracking(Using Kalman Filters and Algorithmic Tricks)
Refer to the CMake project to compile, you need to use PCL and OpenCV.
- Remove the test code and sort out the stripped-down version
- For the problem of point cloud jitter, there is no suitable method to solve it (mainly reflected in tracking)
- Improve the algorithm to the ROS version and remove the dependency on the camera SDK