Local Descriptors without Orientation Normalization to Enhance Landmark Regconition

被引：6

作者：

Dai-Duong Truong ^{[1
]}

Chau-Sang Nguyen Ngoc ^{[1
]}

Vinh-Tiep Nguyen ^{[1
]}

Minh-Triet Tran ^{[1
]}

Anh-Duc Duong ^{[2
]}

机构：

[1] Univ Sci, VNU HCM, Fac Informat Technol, Ho Chi Minh City, Vietnam

[2] VNU HCM, Univ Informat Technol, Ho Chi Minh, Vietnam

来源：

KNOWLEDGE AND SYSTEMS ENGINEERING (KSE 2013), VOL 1 | 2014年 / 244卷

关键词：

D O I：

10.1007/978-3-319-02741-8_34

中图分类号：

TP18 [人工智能理论];

学科分类号：

081104 ; 0812 ; 0835 ; 1405 ;

摘要：

Derive from practical needs, especially in tourism industry; landmark recognition is an interesting and challenging problem on mobile devices. To obtain the robustness, landmarks are described by local features with many levels of invariance among which rotation invariance is commonly considered an important property. We propose to eliminate orientation normalization for local visual descriptors to enhance the accuracy in landmark recognition problem. Our experiments show that with three different widely used descriptors, including SIFT, SURF, and BRISK, our idea can improve the recognition accuracy from 2.3 to 12.6% while reduce the feature extraction time from 2.5 to 11.1%. This suggests a simple yet efficient method to boost the accuracy with different local descriptors with orientation normalization in landmark recognition applications.

引用

页码：401 / 413

页数：13