Computer Vision BASED Moving Object Detection And Tracking

Abstract

To Detect And Track Objects, Some Approaches Of Detecting And Tracking Moving Objects In Stationary Scene Are Presented, Including The Detection Methods Of The Time Domain Differential Method, The Background Differential Method, Optical Flow Method And Tracking Methods Of The Kalman Filter And Feature Optical Flow. Keywords-Moving Object Detection And Tracking; Background Model; Kalman Filter; Optical Flow. Object Tracking Is An Important Task Within The Field Of Computer Vision. It Is A Challenging Problem. Many Difficulties Arises In Tracking The Objects Due To Abrupt Object Motion, Changing Appearance Patterns Of Both The Object And The Scene, Non-Rigid Object Structures, Object-To-Object And Object-To-Scene Occlusions, And Camera Motion. This Paper Selectively Reviews The Research Papers With Regard To Tracking Methods On The Basis Of The Object, Their Motion Representations And All Detailed Descriptions Of Representative Methods In Each Category Examining Their Advantages/Disadvantages.