Probabilistic Visual Learning for Object Representation
IEEE Transactions on Pattern Analysis and Machine Intelligence
Probabilistic Modeling of Local Appearance and Spatial Relationships for Object Recognition
CVPR '98 Proceedings of the IEEE Computer Society Conference on Computer Vision and Pattern Recognition
Face recognition: component-based versus global approaches
Computer Vision and Image Understanding - Special issue on Face recognition
Object Detection Using the Statistics of Parts
International Journal of Computer Vision
Learning to Detect Objects in Images via a Sparse, Part-Based Representation
IEEE Transactions on Pattern Analysis and Machine Intelligence
Scale-invariant shape features for recognition of object categories
CVPR'04 Proceedings of the 2004 IEEE computer society conference on Computer vision and pattern recognition
Visual perception of obstacles and vehicles for platooning
IEEE Transactions on Intelligent Transportation Systems
Recognizing vehicles in infrared images using IMAP parallel vision board
IEEE Transactions on Intelligent Transportation Systems
Preceding vehicle recognition based on learning from sample images
IEEE Transactions on Intelligent Transportation Systems
Face recognition by independent component analysis
IEEE Transactions on Neural Networks
| Hi-index | 0.00 |
This study develops a statistical approach to the automatic detection of vehicles. Compared to traditional approaches, which consider the entire 2-dimensional vehicle region, this study uses three meaningful local features for each vehicle to perform vehicle detection. The proposed approach has a superior tolerance toward wider viewing angles and partial occlusions. Four possible models for vehicle detection are evaluated in the current training and testing processes. For the process of the best model, each local subregion projects into corresponding eigenspace and residual independent basis space with subregion position information. We further simplify the procedure steps of computing the independent component analysis (ICA) in residual space without constructing residual images in order to reduce the computational time. Then the joint probability of projection weight vectors and coefficient vectors of local subregions and positions of local subregions, is used to model the vehicle. Finally, we introduce vector quantization with a new classification method to accelerate the posterior probability calculation.