Lip Feature Points Rough Localization

This program is designed for rough localization of lip feature points in facial images. The algorithm initiates with edge detection applied to lip region images, typically using operators like Canny or Sobel to extract prominent contour boundaries. Following edge extraction, the system analyzes geometric properties including lip shape characteristics, dimensional proportions, and spatial positioning through mathematical morphology operations and contour analysis techniques. Key functions involve calculating curvature extrema, symmetry axes, and proportional relationships between upper and lower lip segments. By processing these geometric attributes, the program achieves preliminary positioning of feature points such as lip corners, cupid's bow peaks, and vermilion borders. This coarse localization provides fundamental data for subsequent precision refinement stages and advanced analysis workflows, enabling more accurate facial expression recognition and lip movement tracking applications. The implementation typically includes threshold parameter configuration for adaptive edge detection and geometric constraint validation to ensure robust performance across varying lighting conditions and lip appearances.