Multilevel Computational Theory of Stereoscopic Visual Image Processing

摘要

This is the final report of a research program that explored the perception of visual forms. The project consisted of psycho-physical studies in which stereoscopically generated, three-dimensional forms were studied in terms of their detectability, discriminability, recognizability, and reconstructability. The project produced a book and 11 articles, all of which spoke to the problem of how humans see forms in space. Two more articles are in progress. This report summarizes the findings of the extensive series of experiments carried out during the program. Results of this study are applicable to a wide range of U.S. Navy interests including computer vision, autonomous vehicles, remotely controlled devices (teleoperators), and displays.