A Two-Stage Object Tracking Method Based on Curvelet Transform and Mean Shift Algorithm

摘要

Traditional mean shift tracking algorithm couldn't track moving objects in cross-scale domain. In this paper, we propose a new two-stage object tracking method combined Curvelet Transform and mean shift algorithm. Our proposed method extracts image features using Curvelet transform, and calculates object location by cross-scale mean shift algorithm. The experimental results demonstrate that the proposed algorithm can effectively track moving objects. Compared with traditional mean shift algorithm, tracking accuracy has been significantly improved.