DocumentCode :
3330446
Title :
Feature-matching based motion prediction for high efficiency video coding in cloud
Author :
Xiang Zhang ; Shiqi Wang ; Shanshe Wang ; Siwei Ma ; Wen Gao
Author_Institution :
Inst. of Digital Media, Peking Univ., Beijing, China
fYear :
2015
fDate :
June 29 2015-July 3 2015
Firstpage :
1
Lastpage :
6
Abstract :
Visual features of images and video frames have become pervasive and maturely developed in extensive research fields such as computer vision and visual search. For realtime retrieval applications, the compact visual features should be transmitted and stored at server side in cloud. These local feature descriptors are characterized by the invariance properties for the variances caused by camera motion, illumination changing, occlusion and different viewpoints. Inspired by these properties, the typical scale-invariant feature transform (SIFT) descriptor is leveraged to improve the video coding efficiency in this work. In particular the predicted motion using SIFT matching is used for merge mode and motion vector prediction (MVP) in the high efficiency video coding (HEVC) standard. A hierarchical motion derivation framework aiming at achieving robust and effective MVP is further proposed. Experimental results have shown that the proposed method can efficiently improve the coding performance according to the accurate feature-matching.
Keywords :
cloud computing; feature extraction; image matching; image motion analysis; video coding; HEVC standard; MVP; SIFT descriptor; camera motion; cloud computing; computer vision; feature-matching based motion prediction; hierarchical motion derivation framework; high efficiency video coding; illumination changing; image frame; invariance property; local feature descriptors; motion vector prediction; occlusion; retrieval applications; scale-invariant feature transform; video frame; visual features; visual search; Encoding; Feature extraction; Image coding; Motion estimation; Robustness; Video coding; Visualization; SIFT; high efficiency video coding; merge mode; motion vector prediction;
fLanguage :
English
Publisher :
ieee
Conference_Titel :
Multimedia & Expo Workshops (ICMEW), 2015 IEEE International Conference on
Conference_Location :
Turin
Type :
conf
DOI :
10.1109/ICMEW.2015.7169778
Filename :
7169778
Link To Document :
بازگشت