Simulation Based Algorithm For Tracking Multi-Moving Object Using Gaussian Filtering With Lucas-Kanade Approach

In this paper, we proposed a method which is used for tracking and detecting multi moving object using Gaussian filtering with Lucas-Kanade approach in a flying bird video. The proposed method is based on output from mean of one frame compared with the result of another frame obtained from optical flow estimator using Gaussian filtering to track the original object. Then after the resulting output of the comparator is fed to perform morphological closing to structuring the elements using strel function. The model is designed such to track multi moving object using Lucas-kanade approach by comparing it's mean in the successive frame of the video with the help of Gaussian filtering. From our knowledge, this proposed method is capable of detecting multi moving object quickly than recently used statistical approaches and also work in noisy environment. (C) 2015 Published by Elsevier B.V.