ildoonet / Ros Openpose
CMU's OpenPose for ROS
Stars: ✭ 81
Projects that are alternatives of or similar to Ros Openpose
Tf Pose Estimation
Deep Pose Estimation implemented using Tensorflow with Custom Architectures for fast inference.
Stars: ✭ 3,856 (+4660.49%)
Mutual labels: robotics, ros, convolutional-neural-networks, pose-estimation
Bonnet
Bonnet: An Open-Source Training and Deployment Framework for Semantic Segmentation in Robotics.
Stars: ✭ 274 (+238.27%)
Mutual labels: robotics, ros, convolutional-neural-networks
Cherry Autonomous Racecar
Implementation of the CNN from End to End Learning for Self-Driving Cars on a Nvidia Jetson TX1 using Tensorflow and ROS
Stars: ✭ 294 (+262.96%)
Mutual labels: robotics, ros, convolutional-neural-networks
Diff drive
ROS nodes for controlling and monitoring a differential drive robot.
Stars: ✭ 77 (-4.94%)
Mutual labels: robotics, ros
Ros Academy For Beginners
中国大学MOOC《机器人操作系统入门》代码示例 ROS tutorial
Stars: ✭ 861 (+962.96%)
Mutual labels: robotics, ros
Pepper plymouth ros
A set of launch files and configuration files for Plymouth University's Pepper robot
Stars: ✭ 22 (-72.84%)
Mutual labels: robotics, ros
Stag
STag: A Stable Fiducial Marker System
Stars: ✭ 75 (-7.41%)
Mutual labels: robotics, pose-estimation
Ev3dev Lang Java
A project to learn Java and create software for Mindstorms Robots using hardware supported by EV3Dev & the LeJOS way.
Stars: ✭ 79 (-2.47%)
Mutual labels: robotics, ros
Joctomap
Java/Android wrapper for Octomap: an octree-based mapping library
Stars: ✭ 11 (-86.42%)
Mutual labels: robotics, ros
Quickmcl
QuickMCL - Monte Carlo localisation for ROS
Stars: ✭ 24 (-70.37%)
Mutual labels: robotics, ros
Champ setup assistant
CHAMP Package Config Generator
Stars: ✭ 24 (-70.37%)
Mutual labels: robotics, ros
Deepseqslam
The Official Deep Learning Framework for Route-based Place Recognition
Stars: ✭ 49 (-39.51%)
Mutual labels: robotics, convolutional-neural-networks
Hrim
An information model for robot hardware. Facilitates interoperability across modules from different robot manufacturers. Built around ROS 2.0
Stars: ✭ 61 (-24.69%)
Mutual labels: robotics, ros
Get Started Ros2
書籍「ROS2ではじめよう 次世代ロボットプログラミング」オンラインリソース
Stars: ✭ 65 (-19.75%)
Mutual labels: robotics, ros
Ros best practices
Best practices, conventions, and tricks for ROS. Do you want to become a robotics master? Then consider graduating or working at the Robotics Systems Lab at ETH in Zürich!
Stars: ✭ 799 (+886.42%)
Mutual labels: robotics, ros
openpose-ros
CMU's Openpose for ROS
Update Information
I re-implemented cmu's openpose in tensorflow with some modifications.
Especially, using Mobilenet's 'Depthwise Separable Convolution', I improved it to run in realtime even in an low-computation embedded deivce or only-cpu environment.
See : https://github.com/ildoonet/tf-openpose
Implementation
- [X] Broadcasting Ros Message
- [X] Humans Pose Estimation
- [ ] Face Landmark
- [ ] Hand Pose Estimation
Installation
Openpose
See Openpose.
After building openpose, set the environment variable OPENPOSE_HOME.
$ export OPENPOSE_HOME='{path/to/your/openpose/home}'
Packages
$ sudo apt-get install ros-indigo-image-common ros-indigo-vision-opencv ros-indigo-video-stream-opencv ros-indigo-image-view
Catkin Build
$ cd src
$ git clone https://github.com/ildoonet/ros-openpose
$ cd ..
$ catkin_make
Message
Parameter
- resolution(string, default: 640x480) : input image resolution
- net_resolution(string, default: 640x480) : network resolution
- num_gpu(int, default: -1) : number of gpus to use. -1 indiciates 'all'.
- num_gpu_start(int, default: 0) : first index of gpu to use.
- model_pose(string, default: COCO) : pose estimation model name provided by openpose.
- no_display(bool, default: false) : if true, it will launch opencv image view to show realtime estimation.
- camera(string, default: /camera/image)
- result_image_topic(string, default: '') : if provided, it will broadcast image with estimated pose.
Broadcast
- /openpose/pose : Msg
BodyParts are stored as indexed Nose = 0 Neck = 1 RShoulder = 2 RElbow = 3 RWrist = 4 LShoulder = 5 LElbow = 6 LWrist = 7 RHip = 8 RKnee = 9 RAnkle = 10 LHip = 11 LKnee = 12 LAnkle = 13 REye = 14 LEye = 15 REar = 16 LEar = 17
Example
Launch File Example
<node name="openpose_ros_node" pkg="openpose_ros_node" type="openpose_ros_node" output="screen" required="true">
<param name="camera" value="/videofile/image_raw" />
<param name="result_image_topic" value="/openpose/image_raw" />
<param name="resolution" value="480x320" />
<param name="face" value="false" />
</node>
Test with Camera/Video Files
$ roslaunch openpose_ros_node videostream.launch video:=0 video_visualize:=true
$ roslaunch openpose_ros_node videostream.launch video:=${filepath} video_visualize:=true
USB Camera Example
See /launch/videostream.
Note that the project description data, including the texts, logos, images, and/or trademarks,
for each open source project belongs to its rightful owner.
If you wish to add or remove any projects, please contact us at [email protected].