Imaging technology (camera) has undergone major development over the past decades. Due to its resolution and reduced size today its integration in computers and mobile phones seems to be natural. In parallel, methods and algorithms related to computer vision as well as their implementation have recently reached a level which made this IT area popular. The application area covers among others medical, production supervision, military, cartographic and object recognition activities.
This thesis has the objective to demonstrate the opportunities given by the Open Source Computer Vision library in relation to a specific and complex task. Its aim is to implement a method which is capable of detecting the position of a Rubik cube side on the camera stream, determining the colours of a given side with a view to have a proprietary model of the application as a result of the process. The complexity of this task enables the demonstration of the main distinct features and difficulties of shape detection.