Please use this identifier to cite or link to this item:
Title: Investigation on mammographic image compression and analysis using wavelets and multiwavelets
Researcher: Ragupathy U S
Guide(s): Tamilarasi, A.
Keywords: Mamographic image compression, wavelets, multiwavelets, Multiwavelet Block Tree Coding
Upload Date: 28-Nov-2013
University: Anna University
Completed Date: 
Abstract: The recent innovation in digital medical imaging techniques requires the development of high performance storage, image transmission and analysis systems. Breast cancer has the highest death incidence rate among women, ranking next to lung cancer. This thesis proposes a solution to the above open problems by investigating the impact of wavelets and multiwavelets on mammographic image compression and analysis. On the one hand, the proposed research work investigates various multiwavelets on mammographic image compression. The first approach in image compression is based on unbalanced multiwavelet transform. The multiwavelets which have irregular basis functions and the filterbank that fails to hold the preservation or annihilation property that requires prefiltering operation are called unbalanced multiwavelets. The second approach in image compression is based on balanced multiwavelet transform. The multiwavelets, which have smooth basis functions and the filterbank that holds both the preservation and annihilation properties possessing similar spectral characteristics, are called balanced multiwavelets. The proposed unbalanced multiwavelet based compression with Multiwavelet Block Tree Coding (MBTC) applied for a set of mammographic images, achieves an average PSNR of 41.16 dB for a bit rate of 0.5 bpp against the existing Set Partitioning In Hierarchical Trees (SPIHT) algorithm. The tests are conducted on mammographic images collected from Mammographic Image Analysis Society (MIAS) database and a few images have been collected from local hospital. The results show that for the testing sample of 30 images, the proposed classification scheme outperforms the scalar wavelet based classification by sensitivity of 2.23% and specificity of 3.4% for original (uncompressed) images. Also it increases the sensitivity by 2.72% and specificity by 8.4% for compressed reconstructed images. This increase in sensitivity and specificity reveals a better performance of the proposed detection scheme. newline newline newline
Pagination: xxvi, 181
Appears in Departments:Faculty of Information and Communication Engineering

Files in This Item:
File Description SizeFormat 
01_title.pdfAttached File49.46 kBAdobe PDFView/Open
02_certificates.pdf1.07 MBAdobe PDFView/Open
03_abstract.pdf18.96 kBAdobe PDFView/Open
04_acknowledgement.pdf14.38 kBAdobe PDFView/Open
05_contents.pdf54.86 kBAdobe PDFView/Open
06_chapter 1.pdf219.71 kBAdobe PDFView/Open
07_chapter 2.pdf50.08 kBAdobe PDFView/Open
08_chapter 3.pdf2.22 MBAdobe PDFView/Open
09_chapter 4.pdf992.8 kBAdobe PDFView/Open
10_chapter 5.pdf786.58 kBAdobe PDFView/Open
11_chapter 6.pdf516.54 kBAdobe PDFView/Open
12_chapter 7.pdf25.5 kBAdobe PDFView/Open
13_appendices 1 to 4.pdf167.86 kBAdobe PDFView/Open
14_references.pdf30.5 kBAdobe PDFView/Open
15_publications.pdf18.19 kBAdobe PDFView/Open
16_vitae.pdf14.71 kBAdobe PDFView/Open

Items in Shodhganga are protected by copyright, with all rights reserved, unless otherwise indicated.