The F-Table: A Data Structure for Rendering Photo-Accurate Images of Faces from Experimentally Acquired Reflectance

• Other Authors: Song, Hui (authoraut)
• Format: Others
• Language: English; English
• Published: Florida State University
• Subjects: Computer science
• Online Access: http://purl.flvc.org/fsu/fd/FSU_migr_etd-1637

My thesis is that a realistic image of a human face can be generated using a computational simulation of light transport. I demonstrate this thesis by acquiring the three-dimensional geometry of a face, measuring the reflectance function from the face, and generating a graphical image that accurately matches a photograph of the face. I also show that the faithful simulation of the physics of light transport is important for synthesizing a realistic image of the face. One important consequence of this work is that it can be used in the future to recognize a face based on a photograph. By rendering realistic images of geometric meshes of faces using accurate re ectance functions and faithful physics, photo-realistic images can be compared pixel by pixel to digitized photographs to search for the closest match between a face in a geometric database and a face in a photograph. === A Dissertation Submitted to the Department of Computer Science in Partial Fulfillment of the Requirements for the Degree of Master of Science. === Summer Semester, 2004. === July 3, 2004. === Photo-Accurate, Photo-Realistic, Exprimental, Recognition, Face, F-Table, Reflectance, Measure, Global Illumination === Includes bibliographical references. === David Banks, Professor Directing Thesis; Anuj Srivastava, Committee Member; Xiuwen Liu, Committee Member.