Masters Theses

Abstract

"A contrast enhancement algorithm is developed for digital mammograms aiming to assist radiologists in discerning early breast cancer easily. The algorithm is based on a Laplacian pyramid framework image processing technique. The mammogram is decomposed into three frequency sub-bands, low, mid, and high frequency sub-band images. The lower sub-band image contains very fine details and higher level contains coarser features. In this method contrast enhancement is achieved from high and mid sub-bands by decomposing the image based on multi-scale Laplacian pyramid and enhance contrast by image processing. Several mapping functions are applied on sub-band images based on experimental analysis. After modifying sub-band images using mapping functions, the final image is derived from reconstruction of the Laplacian images from lower resolution level to upper resolution level. To demonstrate the effectiveness of the algorithm, two mammogram images are analyzed. To validate the algorithm, quantitative measurements are performed. Several existing contrast enhancement techniques are compared with the developed algorithm. Experimental results and quantitative evaluation prove that the proposed algorithm offers improved contrast of digital mammograms"--Abstract, page iv.

Advisor(s)

Lee, Hyoung-Koo

Committee Member(s)

Usman, Shoaib
Kumar, A. S. (Arvind S.)
Moss, Randy Hays, 1953-

Department(s)

Nuclear Engineering and Radiation Science

Degree Name

M.S. in Nuclear Engineering

Sponsor(s)

Nuclear Energy Research Initiative (U.S.)

Publisher

Missouri University of Science and Technology

Publication Date

Fall 2011

Pagination

xi, 49 pages

Note about bibliography

Includes bibliographical references (page 37).

Rights

© 2011 Muhammad Imran Khan Abir, All rights reserved.

Document Type

Thesis - Open Access

File Type

text

Language

English

Subject Headings

Breast -- RadiographyContrast media (Diagnostic imaging)Radiography, Medical -- Digital techniques

Thesis Number

T 9911

Print OCLC #

794671402

Electronic OCLC #

763515749

Download

Share

 
COinS