Object trackers can be broadly classified into two types — feature based and color based. The feature based trackers are scale and illumination invariant whereas color trackers are better at handling occlusions and long term object detection. In this paper we propose a hybrid tracker that uses a feature based Circulant Structure tracker and a color based Mean Shift tracker running in parallel, that can handle scale and illumination changes as well as handle object occlusions and long term object detection in real time. The proposed method is benchmarked against the existing methods on a standard dataset. Further we discuss the application of the proposed method for mobile camera use case where the object selection and camera field of view pose major challenges in object tracking. We show that the proposed method outperforms state of the art short term trackers in sequences captured from mobile camera and is competent in terms of speed and precision on standard dataset. We also show that our failure and detection model for long term tracking perform better than existing TLD framework.