Given a image, how extract (detect) blurred (or focussed) regions? -
Given a image, how extract (detect) blurred (or focussed) regions? -
blurred (focussed) part arbitrary shape (not rectangle)
simple input/output example:
(is valid algorithm, command, software, ...)
thank you!
the variance much higher in focused regions because blurring acts spatial low-pass filter.
but if object shows no variance (uniform object wall or sky) lost , not see differences @ between blurred , focused regions.
therefore question cannot answered every kind of image. it's impossible.
what do:
high pass filter image. (edge detection, absolute differences between neighboring pixels, laplace or local variances, ...) (matlab)
smooth , apply threshold.
apply pre-knowledge imaging optics if existing. illustration knowing focus must round , in in center, ...
image-processing
Comments
Post a Comment