ISSN (Online) : 2319 - 8753 ISSN (Print) : 2347 - 6710 International Journal of Innovative Research in Science, Engineering and Technology Volume 3, Special Issue 3, March 2014 2014 International Conference on Innovations in Engineering and Technology (ICIET’14) On 21 st & 22 nd March Organized by K.L.N. College of Engineering and Technology, Madurai, Tamil Nadu, India M.R. Thansekhar and N. Balaji ;Eds.Ϳ: ICIET’14 2103 Video Distortion Alleviation Using Region Based DT-CWT Fusion M R Mythily 1 , G Rajasekaran 2 1 PG Scholar, Department of information Technology, Mepco Schlenk Engineering College, Tamilnadu, India. 2 Senior Assistant Professor, Department of information Technology, Mepco Schlenk Engineering College, Tamilnadu, India. ABSTRACT- A variety of image restoration methods have been proposed that estimate an improved image by processing a sequence of images. Restoring a scene distorted by atmospheric turbulence is a challenging problem in video surveillance. Image registration enables the geometric alignment of two images and is widely used in various applications in the fields of remote sensing, medical imaging and computer vision. In this paper, we propose a novel method for mitigating the effects of atmospheric distortion on observed images. Region of interest (ROI) for each frame is taken, in order to extract accurate detail about objects behind the distorting layer. A simple and efficient frame selection method is proposed to select informative ROIs, only from good quality frames. Each ROI should be register in order to reduce the distortion. The space varying problem can be solved by image fusion using complex wavelet transform. Finally contrast enhancement is applied. KEYWORDS – Dual tree complex wavelet transform (DTCWT), Image Registration, Image Fusion. I. INTRODUCTION Atmospheric turbulence is a naturally occurring phenomenon that can severely degrade the quality of long range surveillance video footage. Various types of atmospheric distortion can influence the visual quality of video signals during acquisition. Based on temperature variations to reduce the contrast and atmospheric turbulence, due to distortions include fog or haze. When the temperature difference between the ground and the air increases then the thickness of each layer decreases, In strong turbulence, not only scintillation, which produces small-scale intensity fluctuations in the scene [1] and blurring effects are present in the video imagery, but also a shearing effect occurs and is perceived as different parts of objects moving in different directions [2]. To interpret information behind the distorted layer, turbulence effects in the acquired imagery make it extremely difficult. Using various methods, there has been significant research activity attempting to faithfully reconstruct this useful information. In practice, the perfect solution is however impossible, since the problem is ill- posed, despite being simply expressed with a matrix– vector multiplication as in (1). Iobv = D Iidl + ε. (1) Here Iobv and Iidl are vectors containing the observed and ideal images, respectively. Matrix D represents geometric distortion and blur, while ε represents noise. Various approaches have attempted to solve this problem by modeling it as a point spread function (PSF). Where D is considered as a convolution matrix, and then applying deconvolution with an iterative process to estimate Iidl. For the atmospheric distortion case, the PSF is generally unknown, so blind deconvolution is employed [3],[4]. However, the results still exhibit artifacts since the PSF is usually assumed to be space-invariant. Multiple distorted images Restor e mage Frame Selection Image Registrati on Image Debluring Image Fusion Fig. 1 Block diagram of image restoration for atmospheric turbulence.