DocumentCode :
2093943
Title :
A Binary Descriptor Structured on More Spatial Information
Author :
Hui Guobao ; Li Dongbo
Author_Institution :
Sch. of Mech. Eng., Nanjing Univ. of Sci. & Technol., Nanjing, China
fYear :
2013
fDate :
16-18 Nov. 2013
Firstpage :
379
Lastpage :
380
Abstract :
Conventional binary descriptor only uses limited spatial information from original image patch, such as BRIEF, which will result in limited discriminative power. We settle this problem through further excavating feature information and propose a binary descriptor encoding not only intensity comparison information but also intensity order information. Results based on experiments of performance evaluation have shown that the proposed binary descriptor outperforms other binary descriptors under rotation and scale changes.
Keywords :
computer vision; feature extraction; image registration; BRIEF image patch; binary descriptor; discriminative power; feature information; intensity comparison information; intensity order information; performance evaluation; rotation change; scale change; spatial information; Computer vision; Design automation; Detectors; Educational institutions; Image registration; Performance evaluation; Robustness; Computer Vision; binary descriptor; feature match; image patch; keypoint;
fLanguage :
English
Publisher :
ieee
Conference_Titel :
Computer-Aided Design and Computer Graphics (CAD/Graphics), 2013 International Conference on
Conference_Location :
Guangzhou
Type :
conf
DOI :
10.1109/CADGraphics.2013.56
Filename :
6815020
Link To Document :
بازگشت