Robust visual tracking with spatial phase correlation

Abstract

The visual tracker based on phase correlation is always failure, because the response is dirac δ function disturbed by the noise and clutter. Recently, the desired correlated output distribution is adopted in ASEF, which obtains the excellent filtered result. Inspired by the above method, spatial phase correlation is proposed in this paper, which designs the response related with the object spatial position to replace the dirac δ function, which can achieve a robust filter. The phase difference is embedded into the response frequency spectrum to obtain the coarse location of the object. Then the phase saliency is exploited to finely track the object for the excellent performance. To avoid the drifting problem, adaptive template is updated by the peak-sidelobe ratio(PSR), which evaluates the tracking results. Numerical experiments show that the proposed algorithm performs favorably against the state-of-the-art trackers in speed, accuracy and robustness.