MAS.853: Spatial Imaging Systems

Fall Term 2001

Course Information

Instructor: Professor Stephen A. Benton

Office: E15-416, Tel: x3-8145, email: sab@media.mit.edu

Office Hours: Tuesday & Thursday afternoons

(by appointment, via Dina Freedman, x3-0632, olive@media.mit.edu)

Assistant: Wendy Plesniak, E15-422, x3-0375, wjp@media.mit.edu

Hours: by arrangement

Class Schedule: Tuesdays & Thursdays, 10:30 am to 12:00 noon
in Room E15-335 plus lab & project time as scheduled.

Web Site: http://www.media.mit.edu/courses/mas853

Course Outline

1. Introduction: Historical outline of spatial imaging technology. Visual cues to depth. Systems view of spatial image communication.

2. Mysteries of spatial vision: Outline of the human visual system, esp. binocular. Cues to depth perception. Visual illusions and their relation to spatial vision. 3-D machine vision systems.

3. 3-D Photography: Two-eye model system. Visual requirements. Depth and distortion calculations. Space scaling rules. Convergence/accommodation limitations.

4. Stereoscopic Display Technology: Historical & modern evolutions. Imaging optics review. Visual channel separations.

5. Spectacle-based systems: Time-multiplexing (shutter goggles), wavelength multiplexing (additive & subtractive color anaglyphs), polarization multiplexing (projection, Vectographs). 3-D movies.

6. Two-View Autostereoscopic Methods: Specular displays. Swan’s Cube. Raster-screen displays (parallax stereograms, the stereocinema of S. Ivanov). Extensions to multiple views.

7. Lenticular Stereograms: Perspective sampling and multiplexing. Optics of cylinders. Camera systems. Horizontal-parallax-only limitations.

8. Integral Photography: Fly’s-eye lens arrays, the Integram.

9. "Slice-Stacking" Systems: Physical stacking, optical stacking, time multiplexing.

10. Holographic Methods: Basic holography. Holographic analogs: lenticular methods, fly’s-eye lenses, and slice stacking. Projected images.

11. Spatial Information Measures: Comparisons of channel requirements.

12. Included as Needed: Principles of spatial image design. Applications of spatial imaging systems. Optical image formation theory., etc.

13. Special Events: Visiting lecturers, laboratory visits…

Conduct of the Course

The rendering of spatial locations, structures, and shapes is an important aspect of our visual communication repertoire, one that can enhance the speed and accuracy of comprehension of complex data, and one that enhance our emotional engagement with familiar pictorial data. While conventional "flat" graphics can suggest distance and shape relationships, it is the provision of appropriately different perspective views to each of our eyes that provokes the most persuasive, reliable, and accurate perceptions of spatial content, an experience we term "stereopsis." This course will survey a variety of ideas and technologies related to such spatial visualization and imaging, from stereoscopes to holograms, as applied to a wide range of graphical media. It will emphasize current research on the perceptual, technical, and aesthetic bases of effective three-dimensional image communication, with a distinct emphasis on the elimination of viewing aids such as spectacles.

The course will accommodate those wishing to explore spatial imaging techniques as an expressive medium, but is primarily intended as background for advanced imaging and research work. As such, the usual "shop math" level of the course (pre-calculus: algebra, plane geometry, and trigonometry) will occasionally rise to include simple calculus, but without interfering with experimental studies (the problem sets are usually solvable by graphical as well as analytical means). This will make some of the discussions longer and more tedious than absolutely necessary, but hopefully accessible to all. Because of our emphasis on the technology of 3-D displays, a grasp of the elementary concepts of optics is not easy to avoid; instead, these will be introduced and reviewed as they arise, but may require independent study as well.

Although there is a broad literature in the various aspects of "3-D" (see the course reading list), none of the books is a particularly good match to the research orientation of this course. Okoshi’s book is more intellectually responsible than most and will be referred to often, but it is getting a little old (1976, and currently out of print) and not as broad as we need. Thus note-taking in class will be important. A collection of many of the overhead slides used will be maintained in the Optics Lab (E15-441) as the "Spatial Imaging Scrapbook" and will be available for reference.

There will be five problem sets and two quizzes during the term, plus several "laboratory exercises" generally based on field work (such as Polaroid photography). There will be a lab fee for materials of $25. There will not be a final examination, but final projects will be presented during the last week's sessions.

Final Projects

Final projects will become especially important during the second six weeks of the semester. Almost any topic even remotely related to spatial imaging is worthy of consideration. The fundamental requirement is a meaningful depth of treatment, including documented research, and analytical as well as descriptive discussions. Two half-page project proposals for two different projects will be due on Tuesday, October 30, 2001. The experimental part of the projects will be undertaken during the last four weeks of the semester. The projects will be presented as written reports and as oral presentations to the class in two sessions at the end of the semester. during final exam week. Final projects may produce some new technical item (perhaps a demonstration for the lab component of the course) or some new content for an old medium (a visual essay for our image bank, for example) that illustrates some research notion, or a literature research paper on some particular spatial imaging topic. Collaborations between students with substantially different backgrounds (e.g., a designer and an engineer) will be especially warmly received.