计算机科学
Orb(光学)
人工智能
算法
匹配(统计)
特征(语言学)
特征提取
模式识别(心理学)
计算机视觉
Blossom算法
尺度不变特征变换
模板匹配
作者
Jiamin Liu,Jindong Yu,Chudi Wang,Xia Zhang
出处
期刊:International Conference on Computer and Automation Engineering
日期:2018-02-24
卷期号:: 116-119
被引量:1
标识
DOI:10.1145/3192975.3193021
摘要
Feature matching is a key issue to realize a mobile-based augmented reality system. However, the performance of the matching methods is constrained by the processing speed and storage capacity of mobile devices. Considering the real-time requirement of tracking registration in the augmented reality system, a feature matching algorithm based on ORB is proposed, which combines SIFT algorithm with the original ORB algorithm to improve the weakness of the scale invariance on the basis of ensuring the efficiency of the algorithm. Moreover, a Bag-of-Words model is used to enhance the matching speed of feature points. Experimental results show that the algorithm can obtain better results than the original ORB algorithm on the aspects of real-time tracking target, scale invariance and matching time.
科研通智能强力驱动
Strongly Powered by AbleSci AI