Investigation Of Edge Detection Techniques Based On Brain Tumor Images

Rosnan, Murni Nur Athirah (2018) Investigation Of Edge Detection Techniques Based On Brain Tumor Images. Project Report. Universiti Sains Malaysia, Pusat Pengajian Kejuruteraan Elektrik dan Elektronik. (Submitted)

Download (697kB) | Preview


Medical image processing has become an important technique that can visualize the interior of a human body for better diagnosis and extraction of an anatomical structure. Image processing has an advantage which reproduced original data repetitively without any changes that helps radiologist for analysis. Magnetic Resonance Imaging(MRI) is one of the medical imaging modalities that depend on computer technology to create detailed images of the brain. The output image by MRI need to undergo several imaging techniques to extract the important information accurately. In this work, all input MRI brain images are in DICOM format. The images undergo three fundamental steps of edge detection techniques. The edge detection operators used to detect the brain tumor are Robert zero-crossing, Sobel operator, Prewitt operator, Canny operator and modified Canny algorithm. The visual results from each operators are analyzed using quantitative and qualitative measurement. The quantitative parameters used to evaluate the operators performances are PSNR, MSE and SSIM. Based on the quantitative analysis, the new Canny algorithm successfully produced high quality image with less error. However, from visual perspective, Sobel operator produced better edge maps of the brain tumor compared to the Modified Canny algorithm.

Item Type: Monograph (Project Report)
Subjects: T Technology
T Technology > TK Electrical Engineering. Electronics. Nuclear Engineering
Divisions: Kampus Kejuruteraan (Engineering Campus) > Pusat Pengajian Kejuruteraaan Elektrik & Elektronik (School of Electrical & Electronic Engineering) > Monograph
Depositing User: Mr Engku Shahidil Engku Ab Rahman
Date Deposited: 25 Jul 2022 02:49
Last Modified: 25 Jul 2022 02:49

Actions (login required)

View Item View Item