Performance improved GA based statistical computing technique for retinal image segmentation

Retinal vessel segmentation is important for the detection of numerous eye diseases and plays an important role in automatic retinal screening systems. K-nearest neighbor classifier is used to perform a soft segmentation of retinal vessels and is a supervised method. This method produces segmentation by classifying each image pixel as vessel or nonvessel, based on the output of filters and the pixel values with in the neighborhood. Genetic algorithms are powerful tools for K-nearest neighbors classifier optimization. Genetic Algorithm is used to optimize the feature vector by removing both irrelevant and redundant features and finds optimal ones. In this work, GA is used to estimate the K value. The performance of the unoptimised K-nearest neighbor classifier and the genetic optimized K-NN are analysed in terms of segmentation efficiency and convergence time period. Experimental results show superior results for the genetic algorithm based K-NN in terms of the performance measures.