Constrained segmentation of complex models for image-based texture editing

We present a constrained segmentation method for the partitioning of complex textured 3D models. Our method automatically identifies large regions of reduced metric distortion suitable for texture mapping. The boundaries of the regions conform to salient features of the model and respect user-imposed constraints. We illustrate the application of this method to virtual paint restoration through image-based texture editing. © 2004 ACM.