Please use this identifier to cite or link to this item: http://hdl.handle.net/10603/453030
Title: Certain investigations on hybrid image Analysis techniques for diagnosis of Brain tumor
Researcher: Janardhana prabhu, S
Guide(s): Malathi, V
Keywords: Engineering and Technology
Engineering
Engineering Electrical and Electronic
Medical imaging
Brain tumor
University: Anna University
Completed Date: 2021
Abstract: Image processing is a vast domain that processes the images to obtain newlineuseful information and it can be applied in various applications like remote newlinesensing, medical imaging, face detection, fiber defect detection, forgery newlinedetection etc. In image processing, four common processes are used and they are newlinepre-processing, feature extraction, segmentation and classification. The main newlineobjective of the present work is to find the best algorithms to handle different newlinekinds of inputs such as multi-sequence and multi resolution images. The newlineapproaches of each step are selected based on the type of input images. newlineFor pre-processing step, different kinds of pre-processing approaches newlineare needed for different imagings. To remove the salt and pepper and speckle newlinenoises, median filter could be a suitable choice. Thresholding process is another newlinekind for tracing the edges in the images. The selected pre-processing step in this newlinethesis is suitable for extracting the original form of brain MR images. newlineImage segmentation is the process of separating the images into newlinedifferent parts in order to evaluate a specific substance and its margins. Cancer newlinecan be defined as the uncontrolled, unnatural growth and division of the cells in newlinethe body. Occurrence, as a mass, of these unnatural cell growth and division in newlinethe brain tissue is called a brain tumor. Magnetic Resonance Imaging (MRI) is newlinethe most commonly used technique for diagnosing brain tumours. Exact newlineinformation about the affected area is critical for proper treatment. Since MRI newlinediagnosis generates a large amount of data, an automated segmentation newlinetechnique is necessary to obtain precise information of tumor. In the initial work newlineof this thesis, Depth-First Search (DFS) segmentation algorithm based graph newlinetheory has been presented. Based on their proximity in the image, image pixels newlineare arranged into a tree-like structure newline
Pagination: xiv,144p.
URI: http://hdl.handle.net/10603/453030
Appears in Departments:Faculty of Information and Communication Engineering

Files in This Item:
File Description SizeFormat 
01_title.pdfAttached File24.25 kBAdobe PDFView/Open
02_prelim pages.pdf1.61 MBAdobe PDFView/Open
03_content.pdf148.72 kBAdobe PDFView/Open
04_abstract.pdf9.35 kBAdobe PDFView/Open
05_chapter 1.pdf403.78 kBAdobe PDFView/Open
06_chapter 2.pdf350.51 kBAdobe PDFView/Open
07_chapter 3.pdf936.41 kBAdobe PDFView/Open
08_chapter 4.pdf885.06 kBAdobe PDFView/Open
09_chapter 5.pdf1.35 MBAdobe PDFView/Open
10_annexures.pdf154.24 kBAdobe PDFView/Open
80_recommendation.pdf68.51 kBAdobe PDFView/Open
Show full item record


Items in Shodhganga are licensed under Creative Commons Licence Attribution-NonCommercial-ShareAlike 4.0 International (CC BY-NC-SA 4.0).

Altmetric Badge: