LightBlog
Contact at mumbai.academics@gmail.com or 8097636691/9323040215
Responsive Ads Here

Monday, 28 May 2018

A Method of Medical Image Contrast Enhancement Using Two Steps of Contrast Limited Adaptive Histogram Equalization

Abstract— This paper introduced two level contrast limited adaptive histogram equalization for enhancement of mammogram images. Instead of using a single desired histogram, this approach employed two different histograms for contrast limited adaptive histogram equalization (CLAHE) in which one CLAHE follows another CLAHE. Compared to CLAHE, transformation function of the proposed technique is monotonic, which is essential for gray level transformation. This new approach is tested on few images of MIAS database and the improved enhancement performance is compared with adaptive histogram equalization (AHE) and histogram equalization (HE). Keywords— histogram equalization, contrast enhancement, brightness preservation.

INTRODUCTION

In this work we have introduced two level contrast limited adaptive histogram equalization for enhancement of mammogram images. Instead of using a single desired histogram, this approach employed two different histograms for contrast limited adaptive histogram equalization (CLAHE) in which one CLAHE follows another CLAHE. Compared to CLAHE, transformation function of the proposed technique is monotonic, which is essential for gray level transformation. This new approach is tested on few images of MIAS database and the improved enhancement performance is compared with adaptive histogram equalization (AHE) and histogram equalization (HE). The improvement of mammogram is particularly vital in medical imaging since it yields steady hand for analysis reason. In this paper, we build up another technique for upgrading the mammograms by applying Two-Stage contrast limited adaptive histogram equalization (TSCLAHE). There are numerous methods to improve the contrast, which amplifies the power contrast of the mammogram and bringing out more points of interest [1]. Image power dissemination is one of the imperative parameter conversely contrast limited adaptive histogram equalization (CLAHE) since it assumes a noteworthy part in the histogram shape and indicates the desirable histogram [2]. Likewise, CLAHE applies the method on little locales in the image called tiles as opposed to the whole image. Contrast in the tile is upgraded, with the goal that the histogram of the yield area roughly coordinates the uniform distribution. Notwithstanding valuable image insights, the data characteristic in histograms is likewise helpful in image contrast upgrade [3]–[5]. As of late another unsharp masking (UM) plot, called nonlinear UM (NLUM), for mammogram improvement is proposed in [6]. In this new approach, before the uniform appropriation coordinate as in ordinary CLAHE, histogram distribution is utilized to exhibit more concealed inside structure. This will viably help the uniform circulation coordinating by giving more contrast data.

No comments:

Post a Comment