George Legrady: Machine Vision

Camera technology today has become “computational photography”: The integration of computers, sensing systems and algorithmic software embedded into the camera device itself fuses automated aesthetic settings, and post production corrections with enhanced basic operations of recording and storing of the optically captured image. The camera’s standard functions of framing, composition, focus, exposure, are now enhanced, combined with additions of new technological processes that include dynamic tonal range, multi-flash, location through GPS, night vision, and multi-directional optical views, photo montage, panoramic assembly and other such techniques previously unavailable. Vision is further enhanced through sensing systems that go beyond the visible spectrum of the full electro-magnetic range which are applied to recording information in the spatial or time domain. Such methods involve mathematical operations that decompose a non-visible signal into its constituent frequencies, and repositions them into a representation visible to humans.

The lecture will review a number of developments taking place at university research laboratories such as the Camera Culture Lab at MIT; the Computer Vision Lab at Columbia; the Stanford Computer Graphics Lab; Computational Vision Lab at CalTech, and others. Addition discussion will also include engineering research in the areas of the computability of aesthetics and emotions through image analysis to try to describe the rules of “affect”, how the visual syntax in images impress the mind or convey feelings with their intent to formalize and automate systems within the camera that will assist the image-maker in creating images of greater aesthetic and content value.

George Legrady is Director of the Experimental Visualization Lab in the Media Arts & Technology interdisciplinary arts-engineering PhD program at UC Santa Barbara. The lab focuses on new forms of visual representation through the intersections of interactivity with data processing and visualizations. For the past 5 years, he has been a co-principal in a National Science Foundation, IGERT research program in interdisciplinary multimedia research. His pedagogical focus is on synthesizing artistic research and content development with significant engineering processes to achieve tomorrow’s hybrid artist/engineers and engineer/artists. Legrady's recent work focuses on data visualization of public contributed data through large-scale interactive installations that have been featured internationally in Europe, North America and Asia.