Adjacent bin stability evaluating for feature description

Dongdong Nie; Qinyong Ma

doi:10.1117/12.2302609

10 April 2018 Adjacent bin stability evaluating for feature description

Dongdong Nie, Qinyong Ma

Proceedings Volume 10615, Ninth International Conference on Graphic and Image Processing (ICGIP 2017); 106154S (2018) https://doi.org/10.1117/12.2302609
Event: Ninth International Conference on Graphic and Image Processing, 2017, Qingdao, China

Abstract

Recent study improves descriptor performance by accumulating stability votes for all scale pairs to compose the local descriptor. We argue that the stability of a bin depends on the differences across adjacent pairs more than the differences across all scale pairs, and a new local descriptor is composed based on the hypothesis. A series of SIFT descriptors are extracted from multiple scales firstly. Then the difference value of the bin across adjacent scales is calculated, and the stability value of a bin is calculated based on it and accumulated to compose the final descriptor. The performance of the proposed method is evaluated with two popular matching datasets, and compared with other state-of-the-art works. Experimental results show that the proposed method performs satisfactorily.

Citation Download Citation

Dongdong Nie and Qinyong Ma "Adjacent bin stability evaluating for feature description", Proc. SPIE 10615, Ninth International Conference on Graphic and Image Processing (ICGIP 2017), 106154S (10 April 2018); https://doi.org/10.1117/12.2302609

ACCESS THE FULL ARTICLE

INSTITUTIONAL
Select your institution to access the SPIE Digital Library.

SELECT YOUR INSTITUTION

PERSONAL
Sign in with your SPIE account to access your personal subscriptions or to use specific features such as save to my library, sign up for alerts, save searches, etc.

PERSONAL SIGN IN

No SPIE Account? Create one

PURCHASE THIS CONTENT

SUBSCRIBE TO DIGITAL LIBRARY

50 downloads per 1-year subscription

Members: $195

Non-members: $335 ADD TO CART

25 downloads per 1 - year subscription

Members: $145

Non-members: $250 ADD TO CART

PURCHASE SINGLE ARTICLE

Includes PDF, HTML & Video, when available