Bidimensional Empirical Mode Decomposition based Intrinsically Augmented Gamma Correction for Quality Restoration of Textural Images

Abstract

In this paper, texture and illumination improvements for the poorly acquired images are suggested with the proper restoration of images by employing content-dependent decomposition. Being a non-stationary and non-linear two-dimensional digitized signal, any image can be intrinsically decomposed according to its content and hence, content (or behavior) dependent 2-D intrinsic mode functions (2-D IMFs) can be obtained for their individual processing which collectively results into a highly efficient data restoration from the poorly acquired images. In other words, both texture and illumination based improvements can be efficiently entangled in joint space-spatial-frequency domain. Higher mode augmentation, when employed with gamma, corrected illumination boosting in an image-driven and adaptive manner leads to overall quality improvement of the textured data present in the images. In order to validate the necessity of the proposal, a rigorous experimentation is executed by employing the performance evaluation through standard quality measures and comparison with pre-existing recently proposed and highly appreciated quality enhancement approaches.