Paper
24 October 2017 Motion estimation of sequence image based on feature extraction of extended objects
Author Affiliations +
Proceedings Volume 10462, AOPC 2017: Optical Sensing and Imaging Technology and Applications; 1046259 (2017) https://doi.org/10.1117/12.2285765
Event: Applied Optics and Photonics China (AOPC2017), 2017, Beijing, China
Abstract
SIFT feature point extraction algorithm is commonly used in image matching which maintains invariance for scaling, rotation, and brightness changes. Phase correlation algorithm is less dependent on image information with relativity strong noise cancellation performance and high robustness. Based on the traditional phase correlation method, a subpixel accuracy motion estimation method based on interest region is proposed to achieve the high precision localization of the extended object. The paper choose feature blocks centered on feature points extracted from SIFT algorithm. Then the phase correlation operation is performed on the obtained feature blocks and the initial amount of translation is obtained. Then interpolate and implement cubic surface fitting on the neighborhood of the correlation peak, the accurate translation of adjacent images is obtained. The paper simulate the classical pictures applied to image processing. The accuracy of the method is verified by the ideal data.
© (2017) COPYRIGHT Society of Photo-Optical Instrumentation Engineers (SPIE). Downloading of the abstract is permitted for personal use only.
Yazhi Liu and Xinyang Li "Motion estimation of sequence image based on feature extraction of extended objects", Proc. SPIE 10462, AOPC 2017: Optical Sensing and Imaging Technology and Applications, 1046259 (24 October 2017); https://doi.org/10.1117/12.2285765
Advertisement
Advertisement
RIGHTS & PERMISSIONS
Get copyright permission  Get copyright permission on Copyright Marketplace
KEYWORDS
Feature extraction

Motion estimation

Back to Top