Object Detection and Tracking using Zernike Moment

Abstract

In this paper, the cameras capture lot of information that can be highly used in detecting and tracking objects in motion. Though it is challenging, object detection is used to detect object either in each frame or when it first comes in the video. Various methods like frame difference, background subtraction, optical flow can be used for detection. The highest objective of this is to contribute the human operators which help in detecting and tracking doubtful or uncommon actions in the video sequence. The visual observation system needs fast and strong approaches of detecting and tracking the moving objects. Moving objects were detected using canny edge method positively. Moments of different order were extracted using Zernike moments techniques. Analysis was done on all the Zernike moments from z00 to z88. The results show that the number of unique motion in the video and number of unique motion detected by the system is equal with accuracy of 100%.