Real-Time Traffic-Sign Recognition Using Tree Classifiers

Author

Zaklouta, F. ; Stanciulescu, B.

Author_Institution

Robot. Center, MINES ParisTech, Paris, France

Volume

13

Issue

4

fYear

2012

Firstpage

1507

Lastpage

1514

Abstract

Traffic-sign recognition (TSR) is an essential component of a driver assistance system (DAS), providing drivers with safety and precaution information. In this paper, we evaluate the performance of k-d trees, random forests, and support vector machines (SVMs) for traffic-sign classification using different-sized histogram-of-oriented-gradient (HOG) descriptors and distance transforms (DTs). We also use the Fisher´s criterion and random forests for the feature selection to reduce the memory requirements and enhance the performance. We use the German Traffic Sign Recognition Benchmark (GTSRB) data set containing 43 classes and more than 50 000 images.

Keywords

decision trees; image classification; image processing; image recognition; object detection; performance evaluation; road traffic; support vector machines; tree data structures; DAS; DT; Fisher criterion; GTSRB data set; German traffic sign recognition benchmark data set; HOG descriptors; SVM; TSR; different-sized histogram-of-oriented-gradient descriptors; distance transforms; driver assistance system; feature selection; k-d trees; memory requirement reduction; performance evaluation; random forests; real-time traffic-sign recognition; support vector machines; traffic-sign classification; tree classifiers; Image classification; Image processing; Machine learning; Machine vision; Object detection; Pattern recognition; Support vector machines; Advanced driver-assistance systems; image classification; image processing; machine vision; object detection; object recognition; pattern recognition; traffic sign recognition;

fLanguage

English

Journal_Title

Intelligent Transportation Systems, IEEE Transactions on

Publisher

ieee

ISSN

1524-9050

Type

jour

DOI

10.1109/TITS.2012.2225618

Filename

6353220