Adaptive Multi-Stage Image Denoising via Weighted Bilateral Filtering and Curvelet- Based Hard Thresholding

Digital images are frequently impacted through noise, which degrades their appearance andinfluences subsequent analyses. This research presents a multi-stage denoising approach that synergisticallyincorporates spatial and range-based filters with a novel Weighted Bilateral Filter (WBF) and Curvelet-basedHard Thresholding. The approach outperforms standard denoising approaches by accomplishing higher noisereduction and preserving critical image characteristics. Gaussian noise is actively introduced to replicate real-world scenarios, developing a controlled testing environment for accurate evaluation. Standard images such asLena, Barbara, and Cameraman were employed to evaluate the efficacy of the proposed method towardsconventional methods like Median Filtering and Wiener Filtering. The quantitative assessment, employingmetrics, like Peak Signal-to-Noise Ratio (PSNR) and Structural Similarity Index (SSIM), illustrates significantimprovements. The outcome of the proposed techniques has been illustrated regarding PSNR values over 59.28dB and SSIM values reaching 1.0. This approach shows promising applications in surveillance systems, satelliteimage processing, and medical imaging.