Feature extraction: Difference between revisions

Revision as of 22:13, 18 February 2004

Latest revision as of 23:25, 12 August 2023

(202 intermediate revisions by more than 100 users not shown)

Line 1:

#REDIRECT[[Feature engineering]]

'''Feature extraction''' is an area of [[image processing]] which involves using [[algorithm]]s to detect and isolate various desired portions of a digitized image or video stream.

The following is just a list of subject headings to be filled in later

*Template operations & Fourier Transforms

==Low-level==

* [[Edge detection]]

===Curvature===

*Edge direction, changing intensity, autocorrelation

===Image motion===

*Area based, differential approach. [[Optical flow]].

==Shape Based==

===Thresholding===

===Template matching===

===[[Hough transform]]===

*Lines

*Circles/Ellipses

*Arbitrary shapes (Generalized Hough Transform)

==Flexible methods==

*Deformable, parameterized shapes

*Active contours (snakes)