A Robust Model for Traffic Signs Recognition Based on Support Vector Machines
CISP '08 Proceedings of the 2008 Congress on Image and Signal Processing, Vol. 4 - Volume 04
A Method of Fast and Robust for Traffic Sign Recognition
ICIG '09 Proceedings of the 2009 Fifth International Conference on Image and Graphics
Visual sign information extraction and identification by deformable models for intelligent vehicles
IEEE Transactions on Intelligent Transportation Systems
Real-time computation of Zernike moments
IEEE Transactions on Circuits and Systems for Video Technology
| Hi-index | 0.00 |
Road traffic sign detection and recognition system plays a vital role in building an intelligent vehicle system and other safety driven driver assistance systems. Though considerable research has been done in tracking and detecting the traffic signs, still there are slippages in appropriate road traffic sign recognition systems in countries like India. This paper focuses on the implementation of new advanced traffic sign recognition system architecture for the next generation intelligent vehicle systems. The system takes batches of traffic sign images (i.e. training set) as input and creates a repository for training. The system will be trained using features of the input images (in the repository) with the help of the proposed system and its component Tier-1 Classifier called as Identity Modeler. Tier-2 Classifier named as Identity Recognizer, will recognize the image patterns through the feature information obtained during the training phase of the system repository. Further, this paper details the implementation of this generic architecture with 840 images from 28 distinct traffic signs of India (collected manually). Using Zernike Moment algorithm and LIBSVM, the system can produce 78% accuracy (10 -- fold cross validation) and 77% accuracy with new images that were not used for training.