Permalink
Browse files

Readme edits. Env var name is now: MBV_LIBS.

  • Loading branch information...
padeler
padeler committed Oct 30, 2015
1 parent 28af160 commit 0f7be0479d77250b9705dda55ff97d207e695e0e
Showing with 16 additions and 9 deletions.
  1. +12 −5 README.md
  2. +4 −4 runme.sh
@@ -8,6 +8,8 @@ This script uses the Model Based Vision (MBV) libraries created by the Computer

It implements a hand tracker pipeline described first in [Oikonomidis et al: Efficient model-based 3D tracking of hand articulations using Kinect](http://users.ics.forth.gr/~argyros/mypapers/2011_09_bmvc_kinect_hand_tracking.pdf).

<a href="http://www.youtube.com/watch?feature=player_embedded&v=Fxa43qcm1C4" target="_blank" align="right"><img src="http://img.youtube.com/vi/Fxa43qcm1C4/0.jpg" alt="Hand Tracking" width="240" height="180" border="10" /></a>
The software tracks the 3D position, orientation and full articulation of a human hand from markerless visual observations. The developed method:
* estimates the full articulation of a hand (26 DoFs) involved in unconstrained motion
@@ -27,25 +29,31 @@ System requirements:
## Dependencies
The MBV libraries depend on [OpenCV 2.4](http://opencv.org/) and [tbb 4.2](https://www.threadingbuildingblocks.org/).
Additionally the python libs use python2.7 and numpy.
Linux users just install them using apt:
*Linux* users just install them using apt:
```
sudo apt-get install libopencv-dev libtbb2 python-numpy
```
Windows users must first download and install the correct versions from the links above. For the python packages it is suggested to use [anaconda] (https://www.continuum.io/downloads) which comes with many extra packages including numpy.
If you plan to use openni1.x (required for running some of the example scripts):
```
sudo apt-get install libopenni0 libopenni-sensor-primesense0
```
*Windows* users must first download and install the correct library versions from the links above. For the python packages it is suggested to use [anaconda] (https://www.continuum.io/downloads) which comes with many extra packages including numpy.
In order to use the example script you need to install OpenNI1.x to your system. You will also need a Xtion or Kinect RGBD sensor to use it in real time. You can download the OpenNI1.x binaries and Xtion drivers from [here](http://cvrlcode.ics.forth.gr/web_share/OpenNI/OpenNI_SDK/OpenNI_1.x).
## Usage
To run the hand tracker script you need to download for [Ubuntu 14.04 64](http://cvrlcode.ics.forth.gr/files/mbv/v1.0/MBV_PythonAPI_1.0.zip) or [Windows10 64](http://cvrlcode.ics.forth.gr/files/mbv/v1.0/MBV_PythonAPI_1.0.zip) the MBV and HandTracker libraries for your system.
Unzip the package and set an environment variable named *MBV_SDK* to point to the location of the libraries. The runme scripts use the *MBV_SDK* variable to setup the library paths and python paths correctly before running the script.
Unzip the package and set an environment variable named *MBV_LIBS* to point to the location of the libraries. The runme scripts use the *MBV_LIBS* variable to setup the library paths and python paths correctly before running the script.
You are done. Run the `runme.sh` or `runme.bat` script to test the hand tracker.
@@ -56,4 +64,3 @@ Happy Tracking! :)
For questions, comments and any kind of feedback please use the github issues, and the wiki.

@@ -1,9 +1,9 @@
#!/bin/bash
echo "MBV_SDK Environment variable should point to the MBV libraries folder."
echo "MBV_SDK" $MBV_SDK
echo "MBV_LIBS Environment variable should point to the MBV libraries folder."
echo "MBV_LIBS" $MBV_LIBS
echo "Setting LD_LIBRARY_PATH and PYTHONPATH"
export LD_LIBRARY_PATH=$MBV_SDK/libs:$LD_LIBRARY_PATH
export PYTHONPATH=$MBV_SDK/python_libs:$PYTHONPATH
export LD_LIBRARY_PATH=$MBV_LIBS/libs:$LD_LIBRARY_PATH
export PYTHONPATH=$MBV_LIBS/python_libs:$PYTHONPATH

echo "Running the Single Hand Tracker script..."
python src/SingleHandTracking.py

0 comments on commit 0f7be04

Please sign in to comment.