# Title : A framework for hand gesture recognition using machine Learning techniques
# 
# 0. Contributor
#	- Chanseok Kang (2013-20738 // chanseok@csap.snu.ac.kr)
#	- Sangchul Kim  (2013-20760 // stdio@tcs.snu.ac.kr)

# 1. Abstraction
# This program is a framework for hand gesture recognition based on MATLAB
# All tests works properly in MATLAB R2014a & Image Acquisition Toolbox 
# libSVM & Kinect runtime framework are required for recognition

# 2. Instruction
#	1) Collect RightHand Data
#		- Define the filename such as {alphabet_what you want to train}_{#}.png" in "handGestureDataset.m"
#		- When kinect is installed properly, execute "handGestureDataset.m" 
#		- Once kinect detect human, right hand will be catched automatically.
#		- 10 frames will be saved in each 5 seconds
#
#	2) Train the data
#		- Once the dataset folder is defined, execute "trainingAndTest.m"
#		- libSVM is required for training, so we include the libsvm sharedlibrary in windows/linux version
#		- After a while, svmstruct will be generated.
#		- we include test process using svmpredict()
#			- But we remove the print message by recompiling svm for convenience, you cannot see the predict result
#
#	3) Run the Framework
#		- execute "guiRun.m"
#		- once gui is appeared, press start button, then you can test our framework
#		* caution : If you want to turn off the framework, be sure to press the stop button, not close the window
#		- If you cannot execute this framework properly, run "imaqreset" command in MATLAB and retry.

# if you have any question for this framework, please feel free to contact us. Thank you :)