Indian Journal of Science and Technology
DOI: 10.17485/ijst/2015/v8i17/70326
Year: 2015, Volume: 8, Issue: 17, Pages: 1-7
Original Article
Aref Altaf 1* and Arash Raeisi 2
1 Information Technology Engineering (MIS), National Iranian South Oil Company (NISOC); [email protected]
2 Information Technology Management, National Iranian South Oil Company (NISOC); [email protected]
One of the most important and in progressing issue in image processing and machine vision is object tracking. In fact machine tracking is showing the position changing of the object and tracing it in a sequence of video image with a particular aim. Although the history of object tracking returned to military applications, but this object is used in various fields, so it considered by most researchers these days. In this paper an efficient algorithm for object tracking in video images using color and texture features and Support Vector Machine (SVM) is proposed. In the proposed algorithm, object's area is determined by user in the first frame at first, and then a same area is selected as field. If a part of the object is covered completely, Using SVM in the proposed algorithm cause that remains pixels be recognizable. Proposed algorithm is able to track the object yet there is partial covers and reach to suitable solutions.
Keywords: Image Processing, Machine Learning, Object Tracking, Video Images
Subscribe now for latest articles and news.