Curvelet Transform based Enhancements on Retinal Blood Vessel Segmentation Compared with Contourlet Transform

Retinal images are used in several applications, such as ocular fundus operations as well as human recognition. Also, they play important roles in detection of certain diseases in early stages, such as diabetic retinopathy, hypertensive retinopathy, which can be performed by comparison of the states of retinal blood vessels. Intrinsic characteristics of retinal images make the blood vessel detection process difficult. Due to the acquisition process, retinal images often have low gray level contrast and dynamic range. This problem may seriously affect the diagnostic procedure and its results. Several techniques have been used to enhance the image quality. In this work, a new multi-scale method for retinal image contrast enhancement based on the Contourlet transform is presented. The Contourlet transform has better performance in representing edges than wavelets and curvelets, and is therefore well-suited for multi-scale edge enhancement. Contourlet coefficients in corresponding subbands are modified via a nonlinear function for better visualization. Application of this method on images show that the proposed approach outperforms the Curvelet transform, with an encouraging improvement in representing the thin vessels, and might be helpful for vessel segmentation. The segmentation of blood vessels from the enhanced image is performed based on mathematical morphological transform. The morphological operators are used to smoothen the background, allowing vessels, to be seen clearly and to eliminate the non-vessel pixels. The proposed algorithm being simple and easy to implement, is best suited for fast processing applications.