PROJECT TITLE :

Parallel Block Sequential Closed-Form Matting With Fan-Shaped Partitions - 2018

ABSTRACT:

Applying alpha matting to massive pictures may be a difficult task as a result of of its computational complexity. This Project provides a divide and conquer strategy for performing closed-kind matting. The matting problem, outlined for a complete image, is attenuated into systems of linear equations outlined for very tiny blocks of the image. The sizes of the tiny systems are little enough for us to search out solutions efficiently using a direct sparse linear equation system solver. The little systems are solved following a sequential order such that the alpha matte grows from a user scribble. With the block sequential application, matting is performed on fan-formed partitions in parallel on multiple processing cores. Experiments on giant test pictures plus on commonplace benchmark take a look at images show that the proposed parallel block sequential matting provides top quality alpha mattes with smart scalability.