作者: Reza Pourreza-Shahri , Meysam Tavakoli , Nasser Kehtarnavaz
DOI: 10.1016/J.BSPC.2014.02.011
关键词:
摘要: Abstract This paper presents a computationally efficient method for detection of optic nerve head in both color and fluorescein angiography retinal fundus images. It involves Radon transformation multi-overlapping windows within an optimization framework order to achieve computational efficiency as well high rates the presence various structural, color, intensity variations such Three databases STARE, DRIVE, local database have been examined. is shown that this provides while achieving faster processing speeds than existing methods reported comparable rates. For example, rate STARE which most widely used found be 96.3% with time about 3 s per image.