Q+A CENTER

Dr. Heric: While many things come to mind, two stand out. The first is hyperspectral data which was going to be a panacea. That never came to pass. While hyperspectral imagery still has fans, for the most part, it carries a lot of baggage. The other is spectral data in general. We have been using multiband, multispectral imagery for more than two decades; but, most clients desire simple and intuitive solutions. They want straightforward, high spatial resolution images in true color format.

Dr. Levine: Perhaps the most unanticipated development is the pervasiveness of digital imagery in our lives today. In the academic domain, research on practical applications, such as surveillance and biometrics, has become “respectable” where it was slightly frowned upon in the past. Many journals and conferences now are dedicated to applications of current vision theories and it is nearly impossible for a researcher in academia to keep track of the burgeoning imaging field.

Mr. Schatz: It has been a revelation to see the degree to which expectations for ease-of-use and price have been driven by consumer electronics and PCs. We used to think of our products as specialized industrial instrumentation for which it was all right to require significant investments in customer training. That’s no longer the case. Customers expect machine vision products to be as easy to use and inexpensive as consumer electronics—and it is our challenge to live up to those expectations. We also believed that competition was based on who had the faster hardware or more robust software. Now, we strive to build products that are as “insanely great” as the iPod!

AI: Thank you, gentlemen, for your remarks.