Pioneers in this field face definite challenges:
1. Fragmentation, where the image processor can’t connect discontinuous segments of an edge.
2. False edges, where an edge is identified that doesn’t really belong.
3. Focal blur, where a sharp edge is not present because of optical blurring.
Using two cameras oriented stereoscopically helps by giving depth cues that would otherwise have to be inferred from a single image. Above is an artist’s rendering of a NASA Mars rover from Wikipedia Commons.
The prospect of machines that can see for themselves is still on the horizon. Right now, the applications include autonomous rovers and drones and industrial robots, but there are already early prototypes of machines that can assist the blind, drive cars, and take supersmart photos..
----------
Wikipedia on Edge Detection
Previously: Lines and the Brain
This is pretty fascinating, but what will happen to our cabbies and truck drivers when robotic cars do their jobs for them? Those people need jobs just like any of us do. Artificially intelligent machines could put a lot of people out of work in many industries. It may be appropriate to remember a line from Jurassic Park at this time: "...scientists were so preoccupied with whether they could, that they didn't stop to think if they should."
ReplyDeleteThen people just have to find something else to do. ;) Perhaps people in general will have more spare time. It's not like there are loads of hunter/gatherers sitting around with nothing to do these days ever since farming was invented...
ReplyDelete