Region Filling and Object Removal in Images using Criminisi Algorithm

Aryadarsh S; Dhanya Mathew; Neevan R

CALL FOR PAPERS : Feb-2026

Submission Last Date

25-Feb-26

Submit Manuscript Online

FOR AUTHORS

FOR REVIEWERS

ARCHIEVES

DOWNLOADS

Open Access

Region Filling and Object Removal in Images using Criminisi Algorithm

Author(s):

Aryadarsh S , Marian Engineering College, Kazhakkuttom; Dhanya Mathew, Marian Engineering college; Neevan R, IHRD engineering college Attingal, Trivandrum

Keywords:

confidence term, Image Inpainting, isophotes, priority term, source region, target region

Abstract:

Inpainting is the technique which is used for the modification of a particular image in an undetectable form, is as ancient as art itself. There are a number of applications for inpainting, they varies from the restoration of damaged paintings and photographs to the removal/replacement of selected objects. There have been several approaches proposed for the same. In this paper, we present an algorithm that enhances and extends a previously proposed algorithm and it provides faster inpainting. Using our approach, one can use this to inpaint large regions (e.g. to remove an object etc.) as well as it is used to recover small portions (e.g. restore a photograph by removing cracks etc.). The inpainting method is based on the exemplar based approach. The basic idea behind this approach is to find exemplars (i.e. patches) from the image and replace the lost data with it. This technique can be used for the restoration of damaged photographs or damaged film. In contrast with previous approaches, the technique here introduced has an advantage. That is, it does not require the user to specify where the novel information comes from. This is automatically done (and in a fast way), thereby it allows the system to simultaneously fill-in numerous regions which contain completely different structures and surrounding backgrounds. This paper looks forward to improve the algorithm so that the computational complexity is further improved while retaining the quality of inpainting. Here the inpainting algorithm presented here is not meant to be used for inpainting images, but for videos also. We are also investing methods to improve this algorithm to make it more robust so that it can be used with videos in this paper itself. We would like to provide dynamic settings selection, which would allow the end user to obtain a collection of output images from which the finest result can be chosen.