A Novel Approach to Denoising Impulse Noise in Satellite Images Using Adaptive Switching Trimmed Vector Median Filter

Abstract

In this paper, a novel filtering approach was presented to remove the impulse noise in satellite images. Generally, satellite images are affected by speckle, Gaussian, and impulse noise. It is essential to remove the noise present in the image by applying appropriate filters. By attenuating the high-frequency image components and removing noise from the image, essential features are also lost. Due to their exhaustive search mechanism, the widely used median filter and its variants and switching filter methods, which demonstrate good noise reduction, suffer greatly from computational complexity. With these methods, each pixel in a test window is compared against every other pixel to reach a consensus on the accuracy of the test pixel. This paper presents a novel filtering method for denoising satellite images using the Adaptive Switching Trimmed Vector Median Filtering approach, and it has given better performance when compared to the existing algorithms in terms of peak signal-to-noise ratio (PSNR), structure similarity index (SSIM), and root means square error (RMSE).