Comprehensive Toolkit for ASM and AAM Image Analysis

This toolkit combines ASM (Active Shape Models) and AAM (Active Appearance Models) methodologies, foundational computer vision techniques for deformable object recognition and analysis. The package includes core functions for point distribution model training through principal component analysis (PCA), profile matching algorithms for ASM landmark refinement, and appearance parameter optimization for AAM-based texture reconstruction. Implementation features multi-resolution search strategies and wavelet-based texture representations for efficient convergence. The toolbox has been validated through comprehensive testing on facial recognition and medical image segmentation tasks, demonstrating reliable performance in shape constraint enforcement and texture synthesis. To initialize, users must add the toolkit directory to MATLAB's search path using addpath('toolbox_directory') before calling primary functions like trainASM() or fitAAM().