Comparative study on filters with special reference to Pterygium Images

Pterygium is somewhat a wing shaped fibro vascular proliferation that grows around the cornea region of the eye. It occurs normally in people who stay in the areas of UV radiations, apart from this, hot, dusty, dry, windy, smoky environments also contribute for the onset as well as progression of pterygium. Pre-processing aim is to screen the images in raw form and obtain better quality images after suitable transformation which is further used for the purpose of diagnosis and detection. This paper elaborates the comparisons of four nonlinear filters to remove the impulse noise. Pre-processing has been administered on images of pterygium collected from both private clinics and also images available from public dataset. The applied filters performance is measured using performance metrics Peak Signal to Noise Ratio and Mean Square Error. First ordered entropy also tried to assess the structural content of the pre-processed images. The obtained results justifies that, decision based median filter is found to be the better filter both in terms of high Peak Signal to Noise Ratio[PSNR] and low Mean Square Error[MSE]. The entropy values obtained after the application of filters is showed that the decision based median filter showed an average bit content value of 7.4 which is highest among other filters considered.