Design hybrid filter technique for mixed noise reduction from synthetic aperture radar imagery

Abstract

For military and civilian applications, synthetic aperture radar (SAR) imaging is an essential instrument for obtaining images of the Earth's surface. Speckle noise, a form of noise that is multiplicative, generated by conflicting echoes returned from each pixel, has a significant impact on the SAR picture. On SAR pictures, a hybrid filter for mixed noise reduction is used to remove the mixed noises that are present in the data during capture and transmission. Specifically, speckle noise and salt and pepper noises from SAR images. Both are being worked on at the same time to minimize mixed noise in SAR pictures without revealing edges or other features. This study proposes a technique that combines a hybrid filter derived from a statistics filters with nonlinear functions (SFNF). When comparing to mean, median as no adaptive filters, and frost filter, a lee filter, and fuzzy filters as adaptive filters, this hybrid filter produces good results. MATLAB was used to carry out the simulation. To illustrate the filtering technique's performance, quantitative measurements like signal to noise ratio (SNR) procedure, the mean square error (MSE) method, and edge measurement (β) mechanization are used.