# CityEnv_Perception

**Repository Path**: ryontang/CityEnv_Perception

## Basic Information

- **Project Name**: CityEnv_Perception
- **Description**: No description available
- **Primary Language**: Unknown
- **License**: Not specified
- **Default Branch**: master
- **Homepage**: None
- **GVP Project**: No

## Statistics

- **Stars**: 1
- **Forks**: 0
- **Created**: 2020-03-10
- **Last Updated**: 2025-04-11

## Categories & Tags

**Categories**: Uncategorized

**Tags**: None

## README

# CityEnv_Perception
## Introduction

LiDAR perception based on Velodyne32c, there are 6 ros package:
- bouding_box: shape estimate and 3D box visualization.

- depth_cluster: contains segmentation and clustering algorithm which reproducts the work of paper "Efficient Online Segmentation for Sparse 3D Laser Scans" and "Curved-Voxel Clustering for Accurate Segmentation of 3D LiDAR Point Clouds with Real-Time Performance". The code is referred to [PRBonn/depth_clustering](https://github.com/PRBonn/depth_clustering) and [wangx1996/Lidar-Segementation](https://github.com/wangx1996/Lidar-Segementation).
    
- kalman: kalman filter.
    
- I_shape_track: contains shape estimate and l shape feature based target tracking algorithm which reproducts the work of paper "L-Shape Model Switching-Based Precise Motion Tracking". The code is referred to [kostaskonkk/datmo](https://github.com/kostaskonkk/datmo).

- pcl_process: pcl process node which include the ROI extraCTION, RANSAC ground segmentation and euclidean clustering algorithm. The idea comes from the class lesson of "sensor fusion(Udacity)".

- render: pcl viewer.

Thanks to the reference work mentioned above!


The effect of the perception-system is as follows 
<p align="center">
    <img src=".readme/demo1.gif" width="720px" alt=""/>
    <img src=".readme/demo2.gif" width="720px" alt=""/>
</p>

## References
### Ground Segmenters
- [x] **Depth-cluster Segmentation**. PFG–Journal of Photogrammetry, 2017.(In this repository)
    ```bibtex
    @article{Igor2017Efficient,
      title={Efficient Online Segmentation for Sparse 3D Laser Scans},
      author={Igor Bogoslavskyi and Cyrill Stachniss},
      journal={PFG – Journal of Photogrammetry, Remote Sensing and Geoinformation Science},
      year={2017},
    }
    ```

### Non-ground Clusters
- [x] **Cvc-cluster**. IROS, 2019.(In this repository)
    ```bibtex
    @inproceedings{2019Curved,
      title={Curved-Voxel Clustering for Accurate Segmentation of 3D LiDAR Point Clouds with Real-Time Performance},
      author={ Park, Seungcheol  and  Wang, Shuyu  and  Lim, Hunjung  and  Kang, U. },
      booktitle={2019 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS)},
      year={2019},
    }
    ```

### Shape estimation
- [x] **L-shape**. IV, 2017.(In this repository)
    ```bibtex
    @inproceedings{2017Efficient,
      title={Efficient L-shape fitting for vehicle detection using laser scanners},
      author={ Zhang, Xiao  and  Xu, Wenda  and  Dong, Chiyu  and  Dolan, John M. },
      booktitle={2017 IEEE Intelligent Vehicles Symposium (IV)},
      year={2017},
    }
    ```

### Tracking based on 3D object
- [x] **L-Shape Model Switching-Based Precise Motion Tracking**. 2018.(In this repository)
    ```bibtex
    @article{2018L,
      title={L-Shape Model Switching-Based Precise Motion Tracking of Moving Vehicles Using Laser Scanners},
      author={ Kim, Dongchul  and  Jo, Kichun  and  Lee, Minchul  and  Sunwoo, Myoungho },
      journal={IEEE Transactions on Intelligent Transportation Systems},
      volume={19},
      number={2},
      pages={598-612},
      year={2018},
    }
    ```