Earlier work ported from Matlab to C++ for higher performance.
Uses OpenCV (required to compile)
Currently implements the same functionality we had in Matlab, with back/forward gestures. Current keypress code is specific to Mac OS for controlling iBooks, but can be easily changed.
Two colour gesture for zoom in/out is working.
Mouse movement with green is now working. Combine blue with green to trigger a click.
Performance has been improved by using multithreading. One thread captures images from the camera and places them in a queue. Another thread takes these images from the queue and does the processing.
OpenCV documentation suggests using the [HSV colour space] (https://en.wikipedia.org/wiki/HSL_and_HSV) for colour tracking. There are now some on-screen sliders that let you tweak the colour settings to get a good match depending on the condition.
Added voice recognition for "bookmark" to add a bookmark.
Reduced to one preview windows for all colours. https://dl.dropboxusercontent.com/u/12899811/colourss.png Some sample OpenCV code:
http://docs.opencv.org/trunk/df/d9d/tutorial_py_colorspaces.html http://opencv-srf.blogspot.ca/2010/09/object-detection-using-color-seperation.html
Add right clicking
Investigate GPU acceleration. Requires NVidia graphics card with CUDA.