Automated detection of microaneurysms to assess diabetics retinopathy

The microaneurysms (MAs), the first sign of diabetics retinopathy (DR), the first step in automated detection of diabetics retinopathy to prevent vision loss and blindness. The proposed method is done using Randon Transform (RT), Top-hat Transform and multioverlapping window. The project addresses a method for detection of retinal land marks. At the first step optics nerve head (ONH) was detected and masked using RT property. In pre-processing stage Top-hat Transform and combination of filter is used to remove the background. The filter output is compared with performance measure. In main processing, the whole processed image is divided into sub-image and then segmented and masked by applying RT in each sub-image.