Evaluation and Optimization of Texture Extraction and Reconstruction for 3D Eyewear Modeling from Single RGB Images

This study investigates multistage texture extraction and reconstruction techniques for the creation of textured 3D representations of eyeglass frames from single RGB product images. We perform a comparative analysis of established methodologies, assessing their performance specifically within the domain of eyewear products. Our primary focus is on optimizing texture extraction and reconstruction processes to enhance the quality and detail of the resulting 3D meshes. This research aims to advance the accuracy of 3D textured models for eyewear.