https://github.com/CPFL/Autoware 自动驾驶框架比较齐全

Autoware

Integrated open-source software for urban autonomous driving, maintained by Tier IV. The following functions are supported:

  • 3D Localization
  • 3D Mapping
  • Path Planning
  • Path Following
  • Accel/Brake/Steering Control
  • Data Logging
  • Car/Pedestrian/Object Detection
  • Traffic Signal Detection
  • Traffic Light Recognition
  • Lane Detection
  • Object Tracking
  • Sensor Calibration
  • Sensor Fusion
  • Cloud-oriented Maps
  • Connected Automation
  • Smartphone Navigation
  • Software Simulation
  • Virtual Reality

Autoware is protected by BSD License. Please use it on your own responsibility. For safe use, we provide a ROSBAG-based simulation method for those who do not own real autonomous vehicles. In case that you use Autoware with real autonomous vehicles, please formulate safety measures and assessment of risk before field testing.

License

  • New BSD License
    • See LICENSE

Spec Recommendation

  • of CPU cores: 8
  • RAM size: 32GB
  • Storage size: 30GB

Requirements

  • ROS indigo(Ubuntu 14.04) or ROS jade(Ubuntu 15.04)
  • OpenCV 2.4.10 or higher NOTE: As of Octber 2016, Autoware does not support OpenCV 3 or higher. Please use OpenCV 2
  • Qt 5.2.1 or higher
  • CUDA(Optional)
  • FlyCapture2(optional)
  • Armadillo

Please use checkout revision before 2015/OCT/21 if you use Autoware on ROS hydro or Ubuntu 13.04, 13.10.

Install dependencies for Ubuntu 14.04 indigo

% sudo apt-get install ros-indigo-desktop-full ros-indigo-nmea-msgs ros-indigo-nmea-navsat-driver ros-indigo-sound-play ros-indigo-jsk-visualization ros-indigo-grid-map
% sudo apt-get install ros-indigo-controller-manager ros-indigo-ros-control ros-indigo-ros-controllers ros-indigo-gazebo-ros-control ros-indigo-sicktoolbox ros-indigo-sicktoolbox-wrapper ros-indigo-joystick-drivers ros-indigo-novatel-span-driver
% sudo apt-get install libnlopt-dev freeglut3-dev qtbase5-dev libqt5opengl5-dev libssh2-1-dev libarmadillo-dev libpcap-dev gksu libgl1-mesa-dev

NOTE: Please do not install ros-indigo-velodyne-pointcloud package. Please uninstall it if you already installed.

Install dependencies for Ubuntu 15.04 jade

% sudo apt-get install ros-jade-desktop-full ros-jade-nmea-msgs ros-jade-nmea-navsat-driver ros-jade-sound-play
% sudo apt-get install ros-jade-controller-manager ros-jade-ros-control ros-jade-ros-controllers ros-jade-gazebo-ros-control ros-jade-sicktoolbox ros-jade-sicktoolbox-wrapper ros-jade-joystick-drivers ros-jade-novatel-span-driver
% sudo apt-get install libnlopt-dev freeglut3-dev qt5-default libqt5opengl5-dev libssh2-1-dev libarmadillo-dev libpcap-dev gksu libgl1-mesa-dev

NOTE: jsk_visualization and grid_map is not provided in Ubuntu15.04 Jade. Please download it from the following repository and build it by yourself. https://github.com/jsk-ros-pkg/jsk_visualization https://github.com/ethz-asl/grid_map

How to Build

$ cd $HOME
$ git clone https://github.com/CPFL/Autoware.git
$ cd ~/Autoware/ros/src
$ catkin_init_workspace
$ cd ../
$ ./catkin_make_release

###Caffe based object detectors CV based detectors RCNN and SSD nodes are not automatically built.

To build these nodes please follow the respective node's README SSD RCNN

How to Start

$ cd $HOME/Autoware/ros
$ ./run

For Developers

Be careful for changing files under ros/src/sensing/drivers/lidar/packages/velodyne. There is subtree. Original repository is here. If you change those files from this repository, you must use git subtree push. (Please never change and push code if you don't understand git subtree well).

GitFlow, git branching model, is introduced in Autoware repository.

  • When you adding new features, you can branch off your feature branch from develop. you can use the following command. $ git checkout -b feature/[your_branch_name] develop
  • When you find bugs in master branch, you can branch off your hotfix branch from master. you can use the following command. $ git checkout -b hotfix/[your_branch_name] master

More Details here

Documents

See Autoware/docs. As of Aug 2015, we provide only Japanese documents. English documents will be added shortly.

Main Packages

Localization

  • ndt_localizer
  • icp_localizer

Detection

  • lidar_tracker
  • cv_tracker
  • road_wizard

Mission (Global) Planning

  • lane_planner
  • way_planner
  • freespace_planner

Motion (Local) Planning

  • astar_planner
  • lattice_planner
  • dp_planner

Vehicle Control

  • waypoint_follower
  • waypoint_maker

Research Papers for Citation

  1. S. Kato, E. Takeuchi, Y. Ishiguro, Y. Ninomiya, K. Takeda, and T. Hamada. "An Open Approach to Autonomous Vehicles", IEEE Micro, Vol. 35, No. 6, pp. 60-69, 2015.

Demo Videos

Public Road Testing

Instruction Videos

Quick Start

Loading Map Data

Localization with GNSS

Localization without GNSS

Mapping

Detection with SSD

Detection with DPM

Detection with Euclidean Clustering

Traffic Light Recognition

Planning with ROSBAG

Planning with wf_simulator

Planning with Hybrid State A*

Calibration Toolkit

Sample Data

3D map of Moriyama in Nagoya

ROSBAG data of Moriyama driving

Script for generating demo launch files of Moriyama

ROSBAG data for Calibration

IROS 2016 Data

ROSBAG data of Nagoya driving

Contact

Autoware Developers (autoware@googlegroups.com)

To subscribe the Autoware Developers ML,

原文发布于微信公众号 - CreateAMind(createamind)

原文发表时间:2017-03-23

本文参与腾讯云自媒体分享计划,欢迎正在阅读的你也加入,一起分享。

发表于

我来说两句

0 条评论
登录 后参与评论

相关文章

来自专栏智能计算时代

IBM Watson提供的认知计算服务介绍

Cognitive Service Introduction Twitter:@huiwenhan Weibo:@huiwenhan Agenda Wats...

36080
来自专栏CreateAMind

End-to-end Driving via Conditional Imitation Learning

Felipe Codevilla, Matthias Müller, Alexey Dosovitskiy, Antonio López, Vladlen Ko...

10610
来自专栏和蔼的张星的图像处理专栏

8.SSD目标检测之二:制作自己的训练集

最近秋色甚好,一场大风刮散了雾霾,难得几天的好天气,周末回家在大巴上看着高速两旁夕阳照射下黄澄澄的树叶,晕车好像也好了很多。 特地周六赶回来为了周末去拍点素材...

16540
来自专栏CreateAMind

paper:Multi-Level Discovery of Deep Options

10110
来自专栏一棹烟波

OpenGL进行简单的通用计算实例

博主作为OpenGL新手,最近要用OpenGL进行并行的数据计算,突然发现这样的资料还是很少的,大部分资料和参考书都是讲用OpenGL进行渲染的。好不容易找到一...

29870
来自专栏GIS讲堂

ArcGIS Image Server简介以及OL2中的加载

本文讲述Arcgis Image Server相关以及在OL2中如何加载Arcgis Server发布的影像服务。

12320
来自专栏数据结构与算法

BZOJ1007: [HNOI2008]水平可见直线(单调栈)

14810
来自专栏专知

关于写论文说来简单但做起来难的三条建议

A few years ago, we prepared a series of workshops on writing research papers an...

29450
来自专栏CreateAMind

Sim-to-Real: 仿真训练直接迁移到真实机器人

8810
来自专栏小樱的经验随笔

HDU 1874 畅通工程续【Floyd算法实现】

畅通工程续 Time Limit: 3000/1000 MS (Java/Others)    Memory Limit: 32768/32768 K (Jav...

288100

扫码关注云+社区

领取腾讯云代金券