Automatic Microaneurysm Detection of Diabetic Retinopathy in Fundus Images

Abstract

A serious diabetic complication is Diabetic retinopathy (DR), and the initial lesion in diabetic retinopathy is Microaneurysm (MA), so initial MA detection plays a critical role in diabetic retinopathy diagnosis. In this paper, the Joint Dynamic Sparse Representation (JDSR) algorithm is with multiple-channel multiple-feature dictionaries. For MA candidates are first extracted as small image blocks; then the multiple-channel multiple-feature dictionaries for candidate representation are developed. Next, sparse coefficient can be obtained by the JDSR algorithm which can be used for classification. The group sparsity dictionary selection method is also introduced, to form an optimal dictionary.