A New Image Inpainting Approach based on Criminisi Algorithm

Nouho Ouattara,Ghislain Koffi,Georges Laussane,Armand Kodjo

doi:10.14569/ijacsa.2019.0100655

Nouho Ouattara, Ghislain Koffi + Show 2 more

Open Access

https://doi.org/10.14569/ijacsa.2019.0100655

Copy DOI

Abstract

In patch-based inpainting methods, the order of filling the areas to be restored is very important. This filling order is defined by a priority function that integrates two parameters: confidence term and data term. The priority, as initially defined, is negatively affected by the mutual influence of confidence and data terms. In addition, the rapid decrease to zero of confidence term leads the numerical instability of algorithms. Finally, the data term depends only on the central pixel of the patch, without taking into account the influence of neighboring pixels. Our aim in this paper is to propose an algorithm to solve the problems mentioned above. This algorithm is based on a new definition of the priority function, a calculation of the average data term obtained from the elementary data terms in a patch and an update of the confidence term slowing its decrease and avoiding convergence to zero. We evaluated our method by comparing it with algorithms in the literature. The results show that our method provides better results both visually and in terms of the Peak Signal-to-Noise Ratio (PSNR) and Structural SIMilarity index (SSIM).

Highlights

Problems related to image degradation are numerous and occur in several areas
Several authors have shown that the priority function proposed in the Criminisi algorithm using the product of confidence and data terms, negatively influences the filling order and the result of the restoration
The curves obtained from our algorithm effectively show a slower decrease in the confidence term than the Criminisi one and their non-convergence at zero

Summary

Introduction

Problems related to image degradation are numerous and occur in several areas. In the case of medical imaging, especially endoscopic imaging, some highlights may appear as white areas and this is often embarrassing when interpreting the scene [1]-[4]. This is similar to the search for white objects in a foliage. For the correcting of the above-mentioned problems, the technique of digital image inpainting is generally used. Image inpainting is intended to fill in missing parts of an image or to remove information that disturbs the interpretation of a scene. There are mainly two categories of inpainting methods: those based on diffusion processes and those based on exemplars [5]-[7]

Methods

Results

Conclusion