Trust Model for IoT Using Cluster Analysis: A Centralized Approach

At lower noise levels, the majority of filter-based impulse noise removal approaches outperform each other. The purpose of this paper is to design an efficient adaptive pulse coupled neural network (APCNN) technique with improved alpha guided grey wolf optimization (IAgGWO) for the elimination of high-density impulse noise. This noise reduction technique is divided into two stages: the detection of noisy pixels and the replacement of a noisy pixel with a data pixel. The IAgGWO technique is utilised to isolate the optimal values for identifying impulse noisy pixels, and the APCNN filtering technique is used to supplant them. This technique provides more accurate and clean filtered images while preserving critical edge pixel information. To demonstrate the IAgGWO-APCNN strategy's efficacy, various degrees of impulse noise were applied to the image and tested. With PSNR of 42 percent, SSIM of 99 percentand STD of 40 percent on satellite pictures, the suggested noise removal model has proved its unshakable consistency in terms of both qualitative and quantitative assessment.