查看文章

The retinal vasculature structure determines a patient’s health condition. Processes to adequately segment the structure from the fundus images currently do not always produce the best results. Attempt is made to present an effective approach to retinal blood vessel segmentation. There is a high requirement of efficient process for ophthalmologists for early detection of diseases. The output of the segmentation process can be utilized for biometric identification, initial screening of patients and automatic surgical evaluation. A wavelet-based enhancement technique for this process of blood vessel segmentation is presented to obtain more precise results compared to the existing methods. The average accuracy of the process accomplished for the segmentation is close to 97% with a lower time complexity. Publicly available colour retinal image database (DRIVE) is used for the comparative study.