Study on feature extraction algorithm of mobile robot vision SLAM under dynamic illumination

Chunjie Hua; Yanan Yu; Zhongmin Wang

doi:10.1117/12.2544551

18 December 2019 Study on feature extraction algorithm of mobile robot vision SLAM under dynamic illumination

Chunjie Hua, Yanan Yu, Zhongmin Wang

Proceedings Volume 11338, AOPC 2019: Optical Sensing and Imaging Technology; 113381R (2019) https://doi.org/10.1117/12.2544551
Event: Applied Optics and Photonics China (AOPC2019), 2019, Beijing, China

Abstract

When a feature point detection method is used in vision SLAM to match images, environment condition around the robot is uncertain usually. Many influence factors such as rotation, scale, fuzzy as well as illumination in the process of detection have a strong impact on robot's locating and incremental map building. Experiments proved that SIFT, SURF, BRISK, ORB and FREAK have good robustness under normal illumination. However, the illumination is complex in practical applications, and the stability of image features extraction will be affected. Based on a mobile robot vision platform, the speed, repetition rate and matching rate of five feature extraction algorithms above are compared and analyzed with different methods. Under dynamic illumination, the robustness and matching effect of image features with translation, rotation, scale and fuzzy transformations are also compared. Through experimental data analyzing, BRISK features shows better effect under dynamic illumination.

Citation Download Citation

Chunjie Hua, Yanan Yu, and Zhongmin Wang "Study on feature extraction algorithm of mobile robot vision SLAM under dynamic illumination", Proc. SPIE 11338, AOPC 2019: Optical Sensing and Imaging Technology, 113381R (18 December 2019); https://doi.org/10.1117/12.2544551

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