ENSCO is a company that makes track inspection systems for railroads. That’s an interesting application area for machine vision, given the need for high resolution, high-speed acquisition under varying environmental conditions. (For those who are interested in “folding” optical working distance to create compact camera packaging, there’s an interesting figure showing just how the four line scan cameras are incorporated into the system.) It would seem though, from what I read on the ENSCO web site, that currently they only do image acquisition and no analysis.
This must mean that they rely on human visual inspection, albeit performed from the comfort of a cubicle rather than out “in the field.” The problem I see with this is that people, no matter how good their working conditions, still get distracted and make mistakes. Wouldn’t it make sense to have the images reviewed automatically, with anything that appears anomalous being flagged for human review?
Just a thought. (Should I submit a development proposal?)
Subscribe to:
Post Comments (Atom)
Posts
No comments:
Post a Comment