hough

Why not use a Hough transform with each drawn point voting for a center, then average the radius?

A Hough transform is kind of a weird beast; not really your typical image processing problem.

Could you explain how to use a Hough transform to find areas that have text?

Also, even hough Raymond Chen's Old New Thing blog is on microsoft.com, the domain name isn't really thr relevant source info.

Using a Hough transform to find a circle with an unknown radius is expensive.

I got a reasonable result using the Daugman algorithm, but Hough circles seemed to be more erratic.

Even the Hough transform that Hinton is so fond of in this talk.

You can use Hough Circles or the Daugman algorithm for segmentation.

The ones you have chosen are some of the more readily-accessible and, to the uninitiated, can be a delight to learn (esp the Hough Transform, imo).

Reminds me of the Hough transform and the Radon transform.

Detecting the horizon seems trivial enough (hough transform?

I do think they can add more complexity in other areas hough.

They made the statement that, "[t]hough the kitchen is the most common location for a fire in dwellings, kitchen fires are among the least likely to be fatal because cooking usually is attended and occurs during waking hours." This is in reference to their assertion that alarms which trigger from cooking are probably inappropriately placed and should (presumably) be further from the kitchen.

"[T]hough we may not want to be wireheaded now, our wireheaded selves would probably prefer to be wireheaded." The Less Wrong community has been discussing questions related to wireheading for years now.

I’m wondering whether Hinton’s reverse computer graphics/bottom-up hierarchical coordinate frames/Hough transform routing approach [1] is suspsusceptible to this too.