A Voting Margin Approach for the Detection of Retinal Microaneurysms

I. Autio, J.C. Borrás, I. Immonen, P. Jalli, and E. Ukkonen (Finland)


Pattern recognition, Classifier combination, Maximum margin classifiers, Support Vector Machines, Ophthalmol ogy, Diabetic Retinopathy.


We address the problem of detecting microaneurysms in retinal fundus images. Retinal microaneurysms are the ear liest known indicator of diabetic retinopathy, an affliction which may result in blindness. We derive a maximum mar gin classifier capable of utilizing a collection of strong base classifiers, each of which may impose a different similar ity inducing kernel in the input space. Our experiments demonstrate that the resulting classifier is accurate and at least on par with the previous approaches to the problem.

Important Links:

Go Back