Explore MATLAB source code curated for "行人检测" with clean implementations, documentation, and examples.
A MATLAB-based pedestrian detection system integrating HOG (Histogram of Oriented Gradients) and LBP (Local Binary Patterns) algorithms for improved feature extraction and classification accuracy.
Implementation of pedestrian detection algorithms using MATLAB's image processing capabilities
HOG feature extraction function designed for pedestrian detection applications, implementing gradient-based feature computation for shape and texture analysis
This is a complete implementation version including calling files. The entire video can be represented using feature sets computed at different scales and positions. The Hog3D descriptor, proposed by Alexander Klaser, Marcin Marszałek, Cordelia Schmid, and colleagues, extends the HOG concept from static image feature extraction to video sequence feature extraction, achieving excellent results in pedestrian detection within video sequences. The implementation typically involves 3D gradient computation and spatiotemporal block normalization.
Full implementation code for Histogram of Oriented Gradients (HOG) with comprehensive applications in pedestrian detection systems, featuring algorithm explanations and key function descriptions.
A MATLAB implementation of pedestrian detection using HOG (Histogram of Oriented Gradients) features combined with AdaBoost classifier, including extensive training and testing image datasets required for program execution.
A MATLAB-based pedestrian detection code collection utilizing Haar features and SVM, including comprehensive dataset for testing and training.
Implementation of pedestrian detection combining HOG features, LBP features, and SVM classifier with code-level explanations
This is an SVM classifier designed for classifying training samples in pedestrian detection applications, featuring direct MATLAB implementation with built-in support for key machine learning functions.
Hierarchical HOG-based pedestrian detection holds significant research value with practical implementation approaches.