Correlative Analysis of Denoising Methods in Spectral Images Embedded with Different Noises

Abstract

Digital image is one of the primary way of communication in the present digital world. During the acquiring process, the images may become noisy. Noise reduction is a demanding task during the image analysis process without dissimilating the important features. It is the procedure of restoring the original image by discarding unwanted noises and known as Image denoising. The main intention of any noise removal technique is to completely eradicate the noise from the image, such that the resulting image is better than the original image. In this digital era, remote sensing images are widely commercial for environmental monitoring. In this study, a correlative analysis of different noise removal methods using various filters in spectral images is performed. Spectral images are introduced with different types of noise and further filters are applied to denoise the image. The performances of the methods are evaluated using benchmarks: Signal-to-Noise Ratio (SNR) and Peak Signal-to-N oise Ratio (PSNR). Experimental results demonstrate that the SNR and PSNR measures were comparatively higher for all the filters when the image is introduced with Poisson noise.