The focus of my research is to make it less apparent that an image has been edited to the end viewer as a result of cutting or selecting a foreground object from its background as it is typically poorly executed and usually results in the edges of the foreground object being bluntly cut or containing colours from the original background. My aim is to not make it apparent that an image has been edited to the end viewer.
By using a green-screen in the professional set up process that is required, or the magic wand or lasso tool found in most image editing applications, only poor results are received. The aim of my research is to look for new methods to create accurate alpha mattes in real time of an image by providing the user with the appropriate user friendly tools to do so.
The ideas that are presented in image completion (which is image in painting combined with texture synthesis), image matting, matter refinement, GPU programming and user interaction tools to create a new system to accomplish the research goal is of great importance to my plan.
The composition into different scenes of research that has the potential to be extended to video matting in which object move in a sequence of frames can also be looked atMy contact details: firstname.lastname@example.org