I require a python script that can process an image containing dice inside of a fixed round area and count the number of pips showing on on each die face.
Please refer to the attached images for examples of the images that require processing. (More images will be supplied to assist in testing accuracy of recognition)
There is a brief outline of one possible algorithm for achieving this in the comments under this video: https://www.google.com.au/url?sa=t&source=web&rct=j&ei=KDO1U7_ULo2_kQWny4HwCA&url=http://m.youtube.com/watch%3Fv%3D8tYbGweDBCM&cd=4&ved=0CCQQtwIwAw&usg=AFQjCNH7u1djqpzAHYIt6GbcMIq05RYTXA
The script should be able to take a image as a parameter from the command line.
Output should return a JSON object containing the values shown on each die.