Diabetic retinopathy is known to be one of the most frequent and serious eyediseases that typically cause blindness in adults between 20 and 60 years of age.Microaneurysm is one of the most important syndromes in color fundus images.A tool for automatic detection of microaneurysms can significantly reduce theworkload of the ophthalmologists. A multi-stage strategy to screen candidatemicroaneurysms is adopted in this study. The computation time for screening onefundus image of size 1500?1152 is 16.8 seconds. Ten verified microaneurysms aresuccessfully detected, plus 5 putative spots that need to be verified. The resultdemonstrates that the proposed approach is efficient yet effective.
