Please use this identifier to cite or link to this item:
Title: Retinal Image Analysis
Authors: TANG, TAT HOU (鄧達豪)
Department: Department of Computer and Information Science
Faculty: Faculty of Science and Technology
Issue Date: 2015
Citation: TANG, T. H. (2015). Retinal Image Analysis (Outstanding Academic Papers by Students (OAPS)). Retrieved from University of Macau, Outstanding Academic Papers by Students Repository.
Abstract: efficiency nowadays. In the Retinal Image Analysis project, we separate to three parts: The extraction of retinal blood vessels, the optic disc detecting and micro-aneurysms detecting. The extraction of retinal blood vessels: Extraction of retinal blood vessels is an important task of computer-aided diagnosis of medical technology for retinal. There have different filter to extract the vessels, but the Matched Filter is the simplest one to extract the vessel from the retinal image. Though the Retinal image of grayscale that easy to get the blood vessels. The optic disc detecting: For Optic Disc detection (OD) that is a major part of retinal and Optic disc also is an important anatomical feature of retinal. In our project, we are using two main step to implement the detection. That are OD vessel candidate detection and OD vessel candidate matching, The first part OD vessel candidate detection is generate by our retinal blood vessels extraction, but have a little bit different because we want to the OD vessel candidate match become more accurate, so we make the OD Bessel candidate detection become more thinner. Then the OD vessel candidate matching is applied the OD vessel candidate detection by Vessel’s Directional Matched Filter (VDMF). Micro-aneurysms detecting: Micro-aneurysms detecting, it is the challenge part for this project, because it is difficult to detect micro-aneurysms by classifying haemorrhages since it is very small in the image and the area of micro-aneurysms is not much as well. For microaneurysms detection mainly separated to two part: Micro-aneurysms detection, and fine Micro-aneurysms classification. The first part is according apply the filter to detect the micro-aneurysms, and the other one is according a feature table that have 31 feature, to filter out the false micro-aneurysms. More detail and method we will fully discuss of the Chepter3.3.
Instructor: Prof. ZHANG, YIBO
Programme: Bachelor of Science in Computer Science
Appears in Collections:FST OAPS 2015

Files in This Item:
File Description SizeFormat 
OAPS_2015_FST_008.pdf12.37 MBAdobe PDFView/Open

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