Retinal vessel enhancement based on multi-scale top-hat transformation and histogram fitting stretching

• A novel retinal vessel enhancement method is proposed.
• We select structuring element automatically to avoid artificial arbitrariness.
• The stretched gray-scale range for histogram fitting is adaptively determined.
• Our method has good robustness because the parameters are selected adaptively.

Retinal vessels play an important role in the diagnostic procedure of retinopathy. A new retinal vessel enhancement method is proposed in this paper. Firstly, the optimal bright and dim image features of an original retinal image are extracted by a multi-scale top-hat transformation. Then, the retinal image is enhanced preliminarily by adding the optimal bright image features and removing the optimal dim image features. Finally, the preliminarily enhanced image is further processed by linear stretching with histogram Gaussian curve fitting. The experiments results on the DRIVE and STARE databases show that the proposed method improves the contrast and enhances the details of the retinal vessels effectively.