Chronic wounds require accurate morphological characterization and personalized therapeutic devices that conform to patient-specific anatomy. This work presents a parametric modeling workflow based on the Method of Anatomical Features (MAF) for reconstructing anatomical wound surfaces using NURBS and Subdivision (SubD) representations. The pipeline combines deep-learning-based segmentation, curvature-driven MAF point extraction, parametric surface reconstruction, and surface offsetting for personalized cover generation. Experiments show sub-millimeter correspondence between original wound boundaries and reconstructed surfaces, with total processing times below ten seconds per case. We also detail NURBS parameterization, subdivision stopping criteria, and an adaptive MAF thresholding strategy to enhance reproducibility and compatibility with additive manufacturing.