Carnegie Mellon University Libraries | University Archives | Envisioning Robotics | Previous | Next

Image Understanding and Computer Vision

Dr. Kanade Posing for Recording on Stereo Machine, c1997

Courtesy of University Archives

Dr. Kanade’s area of research and expertise has been in the field of computer vision, a field that studies and perfects how computers sense their surroundings. As a doctoral student at Kyoto University, Dr. Kanade conducted the first in-depth study of face recognition and incorporated this work into his thesis. This study produced further analysis of the process of formulating facial features into a computer and allowing the computer to learn, commit to memory, and recognize faces [12pp]. This study of computer understanding of images also led to his 1980 work A Theory of Origami World. In this article, Dr. Kanade discusses programming a computer to understand three dimensional images from a two dimensional world, in other words comprehension of three dimensional drawings of origami on two dimensional paper images.

Through his research in image understanding [37pp], Dr. Kanade has furthered the development of a systematic analysis method for vision comprehension. His work has developed methods of computer understanding of images [27pp] from a more computational approach rather than trying to construct a biological method of image interpretation. This development has led to research in three dimensional recoveries of two dimensional images, such as terrain mapping [10pp] that can be modeled into a three dimensional image for robotic navigation and work on the development of a stereo camera [12pp] that allows for true three dimensional images to be captured. His work on vision carries over into many different fields of robotics and serves as the core for most of his contributions to the field.



J. Dustin Williams, University/Heinz Archivist, Carnegie Mellon University

Carnegie Mellon University Libraries | University Archives | Bibliography | Credits