Method to improve detection of motion with phase correlation

作者: Mark Robertson , Ming-Chang Liu , Toru Kurata , Yutaka Yoneda , Yoshihiro Murakami

DOI:

关键词:

摘要: Phase correlation is an established method for computing motion which relies on the ability to find peaks in a computed phase surface. Two methods improve detect surface are described herein. The first applies theoretically-derived and spatially-varying gain compensates whose amplitudes have been decreased due windowing effects; such effects unavoidable correlation. second uses concepts from matched filters detection of diminished spreading peak energy into surrounding positions Peak allow low-amplitude be properly detected. It possible use only method, or both combined.

参考文章(32)
Mark Robertson, Ming-Chang Liu, Method to estimate segmented motion ,(2010)
Franck Seigneret, Praveen Kolli, Sivayya Ayinala, Prabha Atluri, Nabil Khalifa, Multi-threaded DMA ,(2004)
Klaus Zimmermann, Muhammad Siddiqui, Method for motion estimation ,(2005)
Alexander Sibiryakov, Miroslaw Bober, Dominant motion analysis ,(2005)
Vipin Namboodiri, Mainak Biswas, Keepudi Muni Babu, Method and Apparatus for Periodic Structure Handling for Motion Compensation ,(2009)
Clive Henry Gillard, Motion dependent video signal processing ,(1990)