Hi am sahai i would like to get details on region filling and object removal java source code .. now i am living at Hosur and i last studied in the college/school St.Joseph's college tirupur and now am doing PhDi need help on source code
Post Options:
Posts: 14,118
Threads: 61
Joined: Oct 2014
A new algorithm is proposed to remove large objects from digital images. The challenge is to fill the hole left behind in a visually plausible way. In the past, this problem has been approached by two classes of algorithms: 1) "texture synthesis" algorithms to generate large regions of images from sample textures and 2) inpainting techniques to fill small image gaps . The first has been demonstrated for "textures" -repetition of two-dimensional patterns with some stochasticity; The latter focus on linear "structures" that can be thought of as one-dimensional patterns, such as lines and contours of objects. This article presents a new and efficient algorithm that combines the advantages of these two approaches. First of all we must point out that the texture-based synthesis of specimens contains the essential process required to replicate both texture and structure; However, the success of the propagation of the structure depends to a great extent on the order in which the filling takes place. We propose a best-first algorithm in which confidence in the values of synthesized pixels is propagated in a similar way to the propagation of information in inpainting. Actual colour values are calculated using copy-based synthesis. In this work, the simultaneous propagation of the texture and information structure is achieved by a single efficient algorithm. Computational efficiency is achieved through a block-based sampling process. A number of examples of real and synthetic images demonstrate the effectiveness of our algorithm in removing large objects of occlusion as well as thin scratches. Robustness with respect to the shape of the manually selected target region is also demonstrated. Our results compare favourably with those obtained by existing techniques.