International Journal of Science and Research (IJSR) ISSN (Online): 2319-7064 Index Copernicus Value (2013): 6.14 | Impact Factor (2014): 5.611 Volume 4 Issue 11, November 2015 www.ijsr.net Licensed Under Creative Commons Attribution CC BY Motion Object Detection and Tracking in long Distance Imaging through Turbulent Medium Ashwini Thakare 1 , Swati Nandusekar 2 1 Department of EXTC, Pillai’s College, Mumbai University, Mumbai, India 2 Professor, Department of EXTC, Pillai’s College, Mumbai University, Mumbai, India Abstract: This paper represents, automatic detection and tracking of moving objects such as people and vehicles through relatively long distances are important computer applications, but still are very challenging tasks, because of the effects of the atmospheric path, which induces turbulence caused movements in the scene and also blurs the video sequence. Such degraded videos may increase the miss detection (false negative) and false detection (false positive) rates. There are numerous methods for detecting and tracking moving objects in both static and non static environments but only a few methods having specific features with background movements caused by atmospheric turbulence which may have unique characteristics straight forward background subtraction method is not directly used in turbulence medium so we proposed method for detecting and tracking moving objects significantly degraded for atmospheric effects. Keywords: Atmospheric turbulence, Moving object detection, tracking, adaptive threshold, masking 1. Introduction Videos are actually sequences of images, each of which called a frame, displayed in fast enough frequency so that human eyes can percept the continuity of its content. It is obvious that all image processing techniques can be applied to individual frames. Besides, the contents of two consecutive frames are usually closely related. Object detection in videos involves verifying the presence of an object in image sequences and possibly locating it precisely for recognition. Object tracking is to monitor an objects spatial and temporal changes during a video sequence, including its presence, position, size, shape, etc. This is done by solving the temporal correspondence problem, the problem of matching the target region in successive frames of a sequence of images taken at closely-spaced time intervals. These two processes are closely related because tracking usually starts with detecting objects, while detecting an object repeatedly in subsequent image sequence is often necessary to help and verify tracking. Automatic detection and tracking of moving objects such as people and vehicles through relatively long distances (about two kilometers and above) are important computer vision applications, but yet are very challenging tasks. This is due to the effects of the atmospheric path, which induces turbulence caused movements in the scene (temporal clutter) and also blurs the video sequence. Such degradation sources may increase the miss detection (false negative) and false detection (false positive) rates. Because of the long-distance imaging conditions, the objects in the video frames are affected by turbulence-based movements, and are usually small and blurred, and sometimes their real movement is slow. There are numerous methods for detecting and tracking moving objects in both static and nonstatic environments but only a few methods having specific features with background movements caused by atmospheric turbulence which may have unique characteristics straightforward background subtraction method is not well perform in turbulence medium so we proposed method for detecting and tracking moving objects significantly degraded for atmospheric effects. This method we are applied to the various images of video sequences (Rainy, and windy days) 2. Literature Review Authors Baldani and D. Cozzi [1] proposed a simple and robust method for tracking moving targets in an outdoor scene which may contain background movements (as snow, hail, and swaying leaves). The motion detection is based on a probabilistic method for background subtraction, while the tracking is based on matching the data found by the motion detection algorithm with those found by block matching. Only after having tracked a moving object for several frames (at least three), is it not considered a false target. The algorithm has been applied to various image sequences in different weather conditions (sunny, rainy, snowy, and windy days) Authors O. Barnich and M Van [2] proposed a universal background subtraction algorithm, which takes into account the temporal behavior of each pixel, in order to determine whether that pixel belongs to the background, and adapts the model by choosing randomly which values to substitute from the background model (i.e., instead of replacing the oldest values first). Authors B. Fishbain and L.P. Yaroslavsky proposed a real- time algorithm that compensates for spatio-temporal imag movements due to atmospheric turbulence in video sequences, while keeping the real moving objects in the video unharmed. The method is based on (a) generating a background image using a temporal median filter, (b) detecting moving objects using background subtraction with a single (global) threshold, (c) estimating the motion of each detected object (relative to a reference frame) using optical flow estimation (the estimated motion vectors of an object are considered to be turbulence-induced if their magnitudes are low and their direction variations are high), and (d) stabilizing the video by applying a longer temporal median filter to moving objects with higher probability of being turbulence-induced. Paper ID: NOV151095 274