Pixel Valued Adaptive Single Scale Center Surround Algorithm- An Innovative Approach for Illumination removal

Authors

R. Alarmelu Mangai, Science and Humanities, Anna University, Chennai, IndiaFollow
N. Devarajan, Department of Electrical & Electronics Engineering, Sri Ramakrishna Institute of Technology, Coimbatore, IndiaFollow

Author Country (or Countries)

India

Abstract

This paper proposes an innovative approach for removing the non uniform illumination which is a major artifact in microscopic algal images. Center Surround Algorithm is used for correcting the non uniform illumination where the illumination and reflectance are separated but not completely removed. In this new algorithm the pixel values are replaced based on the geometric mean of the surrounding pixels and applying the Adaptive Single Scale Center Surround Algorithm to remove non uniform illumination. By determining inhomogenity and calculating zeros the illumination is completely removed. By iterating the convolution, the edges are also preserved. Many of the algal species look alike and are precisely identified only based on their features and not purely on their appearance. Non uniform illumination that affects the images is completely removed by the proposed method without affecting the features.

Digital Object Identifier (DOI)

http://dx.doi.org/10.18576/amis/110624

Recommended Citation

Alarmelu Mangai, R. and Devarajan, N. (2017) "Pixel Valued Adaptive Single Scale Center Surround Algorithm- An Innovative Approach for Illumination removal," Applied Mathematics & Information Sciences: Vol. 11: Iss. 6, Article 24.
DOI: http://dx.doi.org/10.18576/amis/110624
Available at: https://digitalcommons.aaru.edu.jo/amis/vol11/iss6/24