MSPB: intelligent SAR despeckling using wavelet thresholding and bilateral filter for big visual radar data restoration and provisioning quality of experience in real-time remote sensing

The main reason behind degradation in Synthetic Aperture Radar (SAR) images is speckle noise which is a critical barrier of enhancing Quality of Experience (QoE) in remote sensing of environment. Speckle noise is multiplicative and behaves as a kind of granular pattern which is more an artifact such that a scattering phenomenon inherently exists in the SAR images. The SAR image despeckling is a technique to suppress the noise and preserve the edges (high-frequency information). This article presents a new Method noise wavelet thresholding-based SAR image despeckling using Pixel neighborhood and Bilateral filter (MSPB) for noise suppression and artifact reduction. In the proposed method, MSPB, wavelet-based thresholding is performed using an intelligent Bayesian thresholding rule followed by the method noise thresholding. The experimental outcomes of the MSPB are visually analyzed over the speckled SAR images. The despeckling results are compared to some conventional and some of the latest despeckling methods in the research topic. The despeckling process is also analyzed by image quality assessment (IQA) metrics including no-reference (e.g., ENL) and similarity-based objective (e.g., SNR) and subjective (e.g., SSIM) metrics to measure the quality of performance. The simulation results on some SAR image big datasets show that MSPB is efficient for offline and real-time applications.