Exploiting Visual Saliency and Bag-of-Words for Road Sign Recognition

Dan XU; Wei XU; Zhenmin TANG; Fan LIU

doi:10.1587/transinf.2013EDP7448

Exploiting Visual Saliency and Bag-of-Words for Road Sign Recognition

Dan XU, Wei XU, Zhenmin TANG, Fan LIU

Full Text Views

0

Cite this

Summary :

In this paper, we propose a novel method for road sign detection and recognition in complex scene real world images. Our algorithm consists of four basic steps. First, we employ a regional contrast based bottom-up visual saliency method to highlight the traffic sign regions, which usually have dominant color contrast against the background. Second, each type of traffic sign has special color distribution, which can be explored by top-down visual saliency to enhance the detection precision and to classify traffic signs into different categories. A bag-of-words (BoW) model and a color name descriptor are employed to compute the special-class distribution. Third, the candidate road sign blobs are extracted from the final saliency map, which are generated by combining the bottom-up and the top-down saliency maps. Last, the color and shape cues are fused in the BoW model to express blobs, and a support vector machine is employed to recognize road signs. Experiments on real world images show a high success rate and a low false hit rate and demonstrate that the proposed framework is applicable to prohibition, warning and obligation signs. Additionally, our method can be applied to achromatic signs without extra processing.

Publication: IEICE TRANSACTIONS on Information Vol.E97-D No.9 pp.2473-2482

Publication Date: 2014/09/01

Publicized

Online ISSN: 1745-1361

DOI: 10.1587/transinf.2013EDP7448

Type of Manuscript: PAPER

Category: Image Recognition, Computer Vision

Authors

Dan XU
  Nanjing University of Science and Technology,Jiangsu University of Science and Technology
Wei XU
  Nanjing University of Science and Technology
Zhenmin TANG
  Nanjing University of Science and Technology
Fan LIU
  Nanjing University of Science and Technology

Keyword

visual saliency, bag-of-words, road sign recognition, color names

Cite this

Copy

Dan XU, Wei XU, Zhenmin TANG, Fan LIU, "Exploiting Visual Saliency and Bag-of-Words for Road Sign Recognition" in IEICE TRANSACTIONS on Information, vol. E97-D, no. 9, pp. 2473-2482, September 2014, doi: 10.1587/transinf.2013EDP7448.
Abstract: In this paper, we propose a novel method for road sign detection and recognition in complex scene real world images. Our algorithm consists of four basic steps. First, we employ a regional contrast based bottom-up visual saliency method to highlight the traffic sign regions, which usually have dominant color contrast against the background. Second, each type of traffic sign has special color distribution, which can be explored by top-down visual saliency to enhance the detection precision and to classify traffic signs into different categories. A bag-of-words (BoW) model and a color name descriptor are employed to compute the special-class distribution. Third, the candidate road sign blobs are extracted from the final saliency map, which are generated by combining the bottom-up and the top-down saliency maps. Last, the color and shape cues are fused in the BoW model to express blobs, and a support vector machine is employed to recognize road signs. Experiments on real world images show a high success rate and a low false hit rate and demonstrate that the proposed framework is applicable to prohibition, warning and obligation signs. Additionally, our method can be applied to achromatic signs without extra processing.
URL: https://global.ieice.org/en_transactions/information/10.1587/transinf.2013EDP7448/_p

Copy

@ARTICLE{e97-d_9_2473,
author={Dan XU, Wei XU, Zhenmin TANG, Fan LIU, },
journal={IEICE TRANSACTIONS on Information},
title={Exploiting Visual Saliency and Bag-of-Words for Road Sign Recognition},
year={2014},
volume={E97-D},
number={9},
pages={2473-2482},
abstract={In this paper, we propose a novel method for road sign detection and recognition in complex scene real world images. Our algorithm consists of four basic steps. First, we employ a regional contrast based bottom-up visual saliency method to highlight the traffic sign regions, which usually have dominant color contrast against the background. Second, each type of traffic sign has special color distribution, which can be explored by top-down visual saliency to enhance the detection precision and to classify traffic signs into different categories. A bag-of-words (BoW) model and a color name descriptor are employed to compute the special-class distribution. Third, the candidate road sign blobs are extracted from the final saliency map, which are generated by combining the bottom-up and the top-down saliency maps. Last, the color and shape cues are fused in the BoW model to express blobs, and a support vector machine is employed to recognize road signs. Experiments on real world images show a high success rate and a low false hit rate and demonstrate that the proposed framework is applicable to prohibition, warning and obligation signs. Additionally, our method can be applied to achromatic signs without extra processing.},
keywords={},
doi={10.1587/transinf.2013EDP7448},
ISSN={1745-1361},
month={September},}

Copy

TY - JOUR
TI - Exploiting Visual Saliency and Bag-of-Words for Road Sign Recognition
T2 - IEICE TRANSACTIONS on Information
SP - 2473
EP - 2482
AU - Dan XU
AU - Wei XU
AU - Zhenmin TANG
AU - Fan LIU
PY - 2014
DO - 10.1587/transinf.2013EDP7448
JO - IEICE TRANSACTIONS on Information
SN - 1745-1361
VL - E97-D
IS - 9
JA - IEICE TRANSACTIONS on Information
Y1 - September 2014
AB - In this paper, we propose a novel method for road sign detection and recognition in complex scene real world images. Our algorithm consists of four basic steps. First, we employ a regional contrast based bottom-up visual saliency method to highlight the traffic sign regions, which usually have dominant color contrast against the background. Second, each type of traffic sign has special color distribution, which can be explored by top-down visual saliency to enhance the detection precision and to classify traffic signs into different categories. A bag-of-words (BoW) model and a color name descriptor are employed to compute the special-class distribution. Third, the candidate road sign blobs are extracted from the final saliency map, which are generated by combining the bottom-up and the top-down saliency maps. Last, the color and shape cues are fused in the BoW model to express blobs, and a support vector machine is employed to recognize road signs. Experiments on real world images show a high success rate and a low false hit rate and demonstrate that the proposed framework is applicable to prohibition, warning and obligation signs. Additionally, our method can be applied to achromatic signs without extra processing.
ER -