Top/project/FingerVision/Software/Installation
English | Japanese
English | Japanese

Menu

  • Top
  • Akihiko Yamaguchi 山口 明彦
  • Project プロジェクト
  • Text テキスト
  • Recent articles 最近の記事
  • Articles 記事一覧 (a to z)
  • Search 検索

FingerVision †

  • Concept
  • Videos
  • Publication
  • Manufacturing
  • Software
  • Integration
  • Team
Access: 1/1458 - Admin

Installation of FingerVision Software

There are two options to install the software package fingervision. One is installing as a part of ay_tools, and the other is standalone installation.

Table of contents 目次
  • Installing ay_tools
  • Standalone Installation
    • Install Dependencies
    • Build the ROS Package
    • Build the Standalone Programs
  • Test the Installation

Installing ay_tools †

For installing fingervision as a part of ay_tools, follow the instruction of: Installation of ay_tools.

↑

Standalone Installation †

For install fingervision without ay_tools, make a clone of the Git repository of fingervision in the work space of rosbuild (such as ~/ros_ws/ay_tools).

$ mkdir ~/ros_ws/ay_tools/
$ cd ~/ros_ws/ay_tools/
$ git clone https://github.com/akihikoy/fingervision.git
↑

Install Dependencies †

  • Install OpenCV. Refer to OpenCV Version.
  • Install Boost C++ Libraries:
    $ sudo apt-get -f install libboost-all-dev libboost-dev
↑

Build the ROS Package †

$ rosmake fingervision
↑

Build the Standalone Programs †

No make system for building them. Refer to the pages of the standalone programs. You can also find a build command in the comment of standalone programs.

↑

Test the Installation †

With dummy data (no actual FingerVision sensors are necessary):

 $ rosrun fingervision stream_file1.sh

Access http://localhost:8080/stream.html and http://localhost:8081/stream.html to see the video streams of dummy data.

This program can be quit by pressing Enter.

Run some standalone programs:

 $ cd (fingervision directory)
 $ cd standalone/
 (Just play the video stream)
 $ ./capture.out fv_3_l.yaml
 (Marker tracking demo; Press c to calibrate)
 $ ./blob_tracker2_test.out fv_3_l.yaml
 (Proximity vision demo)
 $ ./prox_vision_test.out fv_3_l.yaml

Note: In blob_tracker2_test.out and prox_vision_test.out, the calibration is done automatically at the beginning of the program. Press c to calibrate again. Note that the calibration of blob_tracker2_test.out is not good in this setup since typically we use white background during the calibration.

Press q to quit each program.

Launch FingerVision data processing program (ROS version):

 $ roslaunch fingervision fv_3_filtered.launch

You will see four windows. This demo starts fv_core_node for the two video streams. The calibration of the marker tracker is loaded from the files. It also launches a filter to the raw data.

Check the ROS topics (do this in another terminal):

 $ rostopic list
 /fingervision/fv_3_l/blob_moves
 /fingervision/fv_3_l/prox_vision
 /fingervision/fv_3_l/wrench
 /fingervision/fv_3_r/blob_moves
 /fingervision/fv_3_r/prox_vision
 /fingervision/fv_3_r/wrench
 /fingervision/fv_filter1_objinfo
 /fingervision/fv_filter1_wrench

Echo some of them to see the content of the topic:

 $ rostopic echo /fingervision/fv_3_l/wrench

rostopic echo and fingervision fv_3_filtered.launch can be quit by Ctrl+C.





Last-modified:2018-08-18 (Sat) 14:41:25 (1677d)
Link: text/ay_tools/Installation(253d) project/FingerVision/Software(1602d) project/FingerVision/Hackathon-Fall-2018(1658d) text/ay_tools(1670d) project/FingerVision/Software/ROS(1676d) text/ay_tools/TestInstallation(1677d) project/FingerVision/Software/OpenCV(1679d)
Site admin: Akihiko Yamaguchi.
Written by: Akihiko Yamaguchi.
System: PukiWiki 1.5.0. PHP 5.2.17. HTML conversion time: 0.017 sec.