Edge Strength based Fuzzification of Colour Demosaicking Algorithms

Seba Susan; Deepak Aneja

doi:10.14429/dsj.64.4732

Seba Susan Delhi Technological University, New Delhi
Deepak Aneja Delhi Technological University, New Delhi

DOI: https://doi.org/10.14429/dsj.64.4732

Keywords: Colour demosaicking, colour filter array interpolation, edge strength, fuzzy membership

Abstract

In this paper we fuzzify the bilinear colour interpolation and the non-linear colour differential correlative interpolation techniques of colour demosaicking, using the edge strength map of the mosaic image. In doing so we ensure that the edges have less participation in the interpolation thereby minimizing the interpolation error. The result is faster and more accurate colour demosaicking algorithms with application to high resolution images. The overall improvement is quite significant when it comes to colour interpolation of high definition/high resolution images containing a lot of detail or edges as proved by the extensive experimental results on state-of-the-art databases and state-of-the-art comparison.

Defence Science Journal, Vol. 64, No. 1, January 2014, DOI:10.14429/dsj.64.4732

Author Biographies

Seba Susan, Delhi Technological University, New Delhi

Dr Seba Susan received her PhD (Image Processing) from the Indian Institute of Technology ( IIT), Delhi. She is currently Assistant Professor in the Department of Information Technology, Delhi Technological University (DTU), Delhi. Her research areas are statistical pattern recognition and soft computing.

Deepak Aneja, Delhi Technological University, New Delhi

Mr Deepak Aneja is a postgraduate student of the Department of Information Technology, Delhi Technological University, Delhi. His areas of interest include colour image processing and fuzzy systems.