Comparision of Various Noises and Filters for Fundus Images Using Pre-processing Techniques

Denoising an image is the important task in an image processing. Images are noised during the capture and transmission. Image enhancement is the main and primary step in image processing concept. It is used to develop the quality and brightness of an image. Generally fundus images are picked up from various locations, using different cameras. This results in a range of images with diverse quality, in some of the images cannot clearly show or detect the pathologies. The noise removal in the image is the problem based on the noise present in the fundus image. To improve the quality of an image the denoising technique is used to enhance an image. In image processing the removing a noise is a challenge task. Typically we know what type of noise and error on the image, for which in this paper some of the standard noises are used to eliminate noise in an image. In this paper we made an effort to study about the various noises such as Amplifier Noise (Gaussian Noise), Salt and Pepper Noise (Impulsive Noise), Speckle Noise (Multiplicative Noise), Poisson Noise. In this paper various filters (Mean Filter, Median Filter, Gaussian Filter and Wiener Filter) are used to remove noise from fundus images and also compare results with estimation of parametric principles like Mean Square Error (MSE), Normalized Absolute Error (NAE), Normalized Cross Correlations (NK), Peak Signal Noise Ratio (PSNR).The Experiment results shows that wiener and Haar filter arethe best filter for medical and fundus images.