Mean Shift, Mode Seeking, and Clustering
IEEE Transactions on Pattern Analysis and Machine Intelligence
Object Recognition from Local Scale-Invariant Features
ICCV '99 Proceedings of the International Conference on Computer Vision-Volume 2 - Volume 2
Distinctive Image Features from Scale-Invariant Keypoints
International Journal of Computer Vision
Beyond Bags of Features: Spatial Pyramid Matching for Recognizing Natural Scene Categories
CVPR '06 Proceedings of the 2006 IEEE Computer Society Conference on Computer Vision and Pattern Recognition - Volume 2
| Hi-index | 0.00 |
Vehicle tag recognition (VTR) is an important part of intelligent transportation system (ITS) that has many applications. The main parts for this system to recognize the vehicle tags include tag location in video, character extraction from the tag images and single character recognition. As for vehicle tag recognition in video streams, the whole processing time of tag recognition must be no more than 0.1 second in order to achieve real-time performance with the recognition system. In location part, templates match and text check are used to find the tag area quickly. Characters are then segmented using the tag's geometrical characteristic and supplementary with the scan of projection image. Next, a hybrid method combining statistical and structural analysis is used to recognize the single character. Vehicle tag's images must be skew corrected and normalized before recognition because of different distance and direction of camera to the objective vehicles. In the recognition part, statistical methods are used to recognize the input character independently. If the output of the this stage contains characters that belong to confusion sets, then structure analysis method is used to further classify these character images. The system has been used in China, South Africa and USA. It is verified that methods developed are very robust and efficient.