A MODIFIED PATCH PROPAGATION-BASED IMAGE INPAINTING USING PATCH SPARSITY

We present a modified examplar-based inpainting method in the framework of patchsparsity. In the examplar-based algorithms, the unknown blocks of target region are inpainted bythe most similar blocks extracted from the source region, using the available information. Defininga priority term to decide the filling order of missing pixels ensures the connectivity of the objectboundaries. In the exemplar-based patch sparsity approach, a sparse representation of missingpixels is considered to define a new priority term and the unknown pixels of the fill-front patch isinpainted by a sparse combination of the most similar patches. Here, we modify this representationof the priority term and take a measure to compute the similarities between fill-front and candidatepatches. Also, a new definition is proposed for updating the confidence term to illustrate theamount of the reliable information surrounding pixels. Comparative reconstructed test images showthe effectiveness of our proposed approach in providing high quality inpainted images.