Gerónimo D., López Antonio M. Vision-based Pedestrian Protection Systems for Intelligent Vehicles

Springer New York Heidelberg Dordrecht London, 2014.118 p. - ISBN: 978-1-4614-7986-4, ISBN: 978-1-4614-7987-1 (eBook).Automobile’s Impact
Advanced Driver Assistance Systems
Pedestrian Protection Systems
The Role of Computer Vision
Generic Framework
Book OutlineCandidates Generation
D-Based Approaches
D-Based Approaches
Motion-Based Approaches
DiscussionClassification
Preliminary Concepts
Image Descriptors
Pedestrian Models
Pedestrian Classifiers
Holistic Models: Focus on the Features
Templates
Haar Features
Edge Orientation Histograms Features
Histogram of Oriented Gradients Features
Shapelet Features
Local Binary Pattern Features
Dominant Orientation Template Features
Co-Occurrence Features
Covariance Features
Data-Driven Features
Diversified Models: From Features Fusion to Multiple Parts
Combined Features
Classifier Cascades/Ensembles
Multiple Aspects
Multiple (Body) Parts
Multiple Resolutions
Occlusion Handling
Training
Parameters Tuning
Bootstrapping
Data Annotation
Domain Adaptation
DiscussionCompleting the System
Preprocessing
Verification and Refinement
Tracking
Application
Real-Time
DiscussionDatasets and Benchmarking
Datasets
Evaluation Protocols
DiscussionConclusions
State of the Research
Future Challenges