Model free trackers can track the hard-handed objects and analysis of these objects with single bounded box. Model free trackers are playing vital role with simultaneously tracking multiple objects with similar appearance. To solve this we used a new multi-object model free tracker by considering the spatial constraints between the objects. In this paper, we explained an online structured SVM algorithm, which helps to learn the spatial constraints. We demonstrated with experimental results, how the structure-preserving object tracker (SPOT) performance betterments in multi-object tracking and also single object trackers can tracking different parts of the objects occurring at the same time.
Keywords:
SVM Algorithm, Image processing, MATLAB
Cite Article:
"Multi-object model free tracker ", International Journal of Novel Research and Development (www.ijnrd.org), ISSN:2456-4184, Vol.2, Issue 5, page no.148-155, May-2017, Available :http://www.ijnrd.org/papers/IJNRD1705032.pdf
Facebook Twitter Google+ Pinterest LinkedIn