Title :
Enhancement of sputum cytology images through recursive mean separate histogram equalization and SVM classification
Author :
Shajy, L. ; Smitha, P. ; Marichami, P.
Author_Institution :
Dept. of Computer Science and Engineering, College of Engineering Karunagappally, India
Abstract :
Contrast enhancement is one of the important steps in image processing. Enhancement process has a vital role in medical image processing. Histogram Equalization (HE) plays the major role in enhancement process. HE is simple and effective method in contrast enhancement. The conventional HE enhancement process outputted an excessive contrast result. Which leads to poor classification result, especially in medical image processing. In this paper we discussed about various HE methods for the enhancement of sputum cytology images. Our ultimate aim is, to develop an efficient algorithm to detect lung cancer at early stage. The challenging problem, we faced, in this work is to find out a proper algorithm for the enhancement of sputum cytology images. Here we consider some famous HE algorithm for the enhancement of sputum cytology images. The Recursive Mean Separate Histogram Equalization Method (RMSHE) gives better result in sputum cytology image enhancement.
Keywords :
Brightness; Cancer; Classification algorithms; Histograms; Image segmentation; Lungs; Support vector machines; classification; enhancement; feature extraction; segmentation;
Conference_Titel :
Computational Intelligence and Computing Research (ICCIC), 2014 IEEE International Conference on
Print_ISBN :
978-1-4799-3974-9
DOI :
10.1109/ICCIC.2014.7238495