Resbee Publishers Multimedia Research (MR) Received 27 May, Revised 25 June, Accepted 3 July Resbee Publishers Vol.3 No.3 2020 29 Adaptive Filter using Improved Pigeon Inspired Optimization Algorithm for Satellite Image Denoising Thomas Thangam Lecturer, International Maritime College of Oman, Sohar, Sultanate of Oman thomasthangam13@gmail.com Abstract: Satellite imaging is a current development in image processing; however, it faces a lot of challenges because of the environmental factors. For denoising, state-of-the-art method has developed some filters like the hyperspectral satellite images, which is not effectual. Moreover, this paper proposed an adaptive filter using the assist of an optimization approach for the satellite image denoising. The developed adaptive filter performs the image denoising via noise correction, noise identification, and image enhancement. In the satellite image by transforming the image to a binary image, the type-2 fuzzy filter recognizes the noisy pixels which are passed via the adaptive non-local mean filter for the noise correction. Subsequently, the kernel-based interpolation scheme performs the image enhancement, which is performed through the developed improved Pigeon optimization algorithm (IPOA). The whole experimentation of the developed denoising system is performed taking into consideration by satellite images from standard databases. It is obvious that the developed adaptive filter with the developed improved Pigeon optimization algorithm has enhanced performance with the PSNR values from the outcomes. Keywords: Image Processing; Denoising; Fuzzy Filter; Kernel; Optimization Algorithm Nomenclature Abbreviations Descriptions ANLMF Adaptive Non-Local Mean Filtering PSNR Peak Signal Noise Ratio ACS Adaptive Cuckoo Search BDCS Bandelet Denoising Compressed Sensing PSO Particle Swarm Optimization IBT Iterative Bandelet Thresholding CS Cuckoo Search SSIM Structural Similarity Index Measure NLMF Non-Local Means Filter GSTV-SC Group Sparse Total Variation with Stepsize Constraints GA Genetic Algorithm CNN Convolutional Neural Network NDVI Normalized Difference Vegetation Index WF Weiner Filter DA Dragonfly Algorithm NIR Near-Infrared Band PIO Pigeon Inspired Optimization 1. Introduction In the present time, satellite images encompass an overabundance of applications predominantly in the areas of weather forecasting, oceanographic studies, forestry, and agriculture, planning and intelligence and so on. The edges or the high-frequency components, in attendance in those images comprise a very important part of the information. Regrettably, owing to the image acquisition lacking and transmission systems, the images acquire gets worsen by noise. Consequently, there arise awe-inspiring needs to extend a denoising approach for the noise removal from degraded satellite images. For satellite images, researchers developed extensive diversity of denoising systems. Although, the improvement of a capable