Research Artcle Journal of Advanced Research in Geo Sciences & Remote Sensing (ISSN: 2455-3190) Copyright (c) 2019: Advanced Research Publicatons Journal of Advanced Research in Geo Sciences & Remote Sensing Volume 6, Issue 3&4 - 2019, Pg. No. 36-38 Peer Reviewed Journal I N F O A B S T R A C T Remote Sensing Satellites are imaging the earth and providing diferent types of imagery in terms of volume, variety, velocity and variability. The spatal resoluton is an important quality parameter and critcal botleneck in various remote sensing applicatons. In this work, we used mult-fractal based reconstructon to increase the spatal resoluton of the remote sensing image. In multfractal analysis, image is treated as a nontrivial combinaton of a number of fractals. Multfractal characteristcs of the low resoluton image are extracted to compute the informaton transfer functon and noise parameters. We generated an enhanced resoluton image using low resoluton image by a fractal based denoising and downscaling method. The reconstructed super resoluton image is validated with original high resoluton image through quality parameters like correlaton coefcient and Structural Similarity Index (SSIM). Keywords: Image Reconstructon, Multfractal Analysis, Informaton Transfer Functon, Downscaling, Super Resoluton Corresponding Author: E Venkateswarlu, National Remote Sensing Centre, Indian Space Research Organisaton, Hyderabad, India. E-mail Id: venkateswarlu_e@nrsc.gov.in Orcid Id: htps://orcid.org/0000-0001-5860-6687 How to cite this artcle: Sneha C, Venkateswarlu E, Nair T. Reconstructon of Satellite Remote Sensing Images using Multfractal Analysis. J Adv Res Geo Sci Rem Sens 2019; 6(3&4): 36-38. Date of Submission: 2019-07-30 Date of Acceptance: 2019-11-15 Reconstruction of Satellite Remote Sensing Images using Multifractal Analysis Ch Sneha 1 , E Venkateswarlu 2 , Thara Nair 3 , GP Swamy 4 , Vinod M Bothale 5 1 S.R.K.R. Engineering College, Bhimavaram, Andhra Pradesh 534204, India. 2,3,4,5 Natonal Remote Sensing Centre, Indian Space Research Organisaton, Hyderabad, India. Introduction Satellite Remote Sensing is an important source to observe the earth, providing various types of imagery in terms of volume, variety, velocity and variability. The spatal resoluton is an important quality parameter and critcal botleneck in various remote sensing applicatons. Super Resoluton (SR) reconstructon is an important method to generate enhanced spatal resoluton image from one or more Low Resoluton (LR) images. 1 The super resoluton reconstructon method can involve single image or mult images based on the number of low resoluton images. The mult image SR reconstructon aims to combine complementary informaton from diferent images based on sub-pixel shifs. The single image SR reconstructon method focuses on extractng relatonships among neighborhood pixels. 2,3 Fractal theory is very efcient method to depict chaotc, erratc, natural phenomena concertzed by Mandelbrot and also considered as straightorward method to analyze extreme variability over a wide range of scales. 4 Here, we used multfractal reconstructon method to enhance the image resoluton. The self similarity in the image is helpful to estmate the informaton in higher and lower scales. We frst search for multfractal characteristcs in the images and then estmate informaton transfer functon and noise in the low resoluton image. Later, we use a fractal based denoising and downscaling method to generate a noise free enhanced resoluton image. This method is useful for any images with multfractal characteristcs along with remote sensing imagery. 5 A multfractal system is a generalizaton of a fractal system which needs contnuous spectrum of exponents. Methodology The relaton between high resoluton image (H) and the low resoluton image (L) is represented by L = H * s + e (1)