<![CDATA[Puspose: This study aims to develop a modified Conditional Random Recursive Central Tendency Filter (CRRCTF) with a fixed-size window to effectively address the challenges of salt-and-pepper impulse noise, particularly under high noise density conditions, while preserving critical image details.Method: The proposed approach is divided into three main phases: (1) detecting noisy pixels through a statistical thresholding mechanism, (2) applying pre-edge filtering to retain edge details, and (3) restoring noisy pixels using a central tendency-based recursive process. Quantitative evaluations were conducted using standard image datasets as well as SAR and optical satellite images to assess the method's robustness.Findings: Experimental results demonstrate the superior performance of the proposed filter, achieving average PSNR and SSIM values of 31.88 and 0.896, respectively, across noise densities ranging from 10 percent to 90 percent. For satellite images, the method achieved PSNR and SSIM values of 29.37 and 0.8096 for SAR images and 19.51 and 0.605 for optical images at an 80 percent noise density.Significance: The proposed CRRCTF method outperforms existing denoising algorithms in terms of image restoration quality, particularly under extreme noise conditions, making it a valuable tool for image preprocessing applications in both research and practical scenarios.]]>
Copyrights © 2025
