object_tracking0

ROS Package for Object Detecton/Tracking

Please notice this repository is still in progress.

Intro

This package includes Ground Removal, Object Clustering, Bounding Box, IMM-UKF-JPDAF, Track Management and Object Classification for 3D-LIDAR multi object tracking. The idea is mainly come from this paper.

Setup

Frameworks and Packages

Make sure you have the following is installed:

• ROS Kinetic
• PCL 1.7.2
• Open CV 3.2【PCL自带opencv，不用安装】

Dataset

• Download the Kitti Raw data.

// 地址已失效

wget http://kitti.is.tue.mpg.de/kitti/raw_data/2011_09_26_drive_0002/2011_09_26_drive_0005_sync.zip
wget http://kitti.is.tue.mpg.de/kitti/raw_data/2011_09_26_calib.zip

// 使用下面下载
https://s3.eu-central-1.amazonaws.com/avg-kitti/raw_data/2011_09_26_calib.zip
https://s3.eu-central-1.amazonaws.com/avg-kitti/raw_data/2011_09_26_drive_0005/2011_09_26_drive_0005_sync.zip

• Convert raw data to rosbag by using the tool made by tomas789. This is his repository.

Start

PLEASE make sure you load the files, src/ego_velo.txt and src/ego_yaw.txt in src/imm_ukf_jpda.cpp l68, l69

Terminal 1

roscore

Terminal 2

rosbag play ~/data/KittiRawdata/2011_09_26_drive_0005_sync/kitti_2011_09_26_drive_0005_synced.bag 

# 下面`--loop`循环paly不推荐加，tracking和上一帧有关，误差越来越大
rosbag play ~/data/KittiRawdata/2011_09_26_drive_0005_sync/kitti_2011_09_26_drive_0005_synced.bag --loop

Terminal 3

rviz

Terminal 4

rosrun object_tracking0 main input:=/kitti/velo/pointcloud

Result

Youtube Clip