Comparison of Machine Learning Algorithms in Icon Recognition

The main aim of this project consists in the study and comparison of Machine Learning
algorithms using them for icon recognition in Graphic Code. Graphic Code is a new Machine Readable Code recently developed by the VisTeam (Institue of Systems and Robotics) that combines the capacity of sending a big amount of information with good aesthetic, improving the capacity when compared with QRcodes, for instance. The methods used for the comparison are the Support Vector Machine and the Convolutional Neural Network. These two algorithms have been chosen because of their ability to recognise different figures and separate them correctly.