Wolfram Language & System 10.4 (2016)|Legacy Documentation

This is documentation for an earlier version of the Wolfram Language.
BUILT-IN WOLFRAM LANGUAGE SYMBOL

ImageCorners

ImageCorners[image]
finds corners in image and returns their coordinates.

ImageCorners[image,r]
finds corners at a pixel range r.

ImageCorners[image,r,t]
uses a threshold t for selecting corners.

ImageCorners[image,r,t,d]
returns corners that are at least d+1 pixels apart.

Details and OptionsDetails and Options

• ImageCorners[image] finds corners in image and returns their image coordinates as a list of the form , where the are ranked by decreasing measure of corner strength.
• ImageCorners[image] is equivalent to ImageCorners[image,2,0,1].
• The following options can be given:
•  MaxFeatureDisplacement 0 maximum allowed corner position refinement MaxFeatures All maximum number of corners to be returned Method Automatic method used in corner filtering
• By default, ImageCorners will return positions at the center of pixels. The positions can be refined to subpixel accuracy. With , a range-p neighborhood is used to compute a refinement of the corner position that may be moved by up to p pixels in either direction.
• By default, corners are detected using a non-maximum suppression on the corner metric computed using CornerFilter. With Method->{method,"StrengthFraction"->f}, only local maxima whose neighbors are less than a fraction f from the center pixel are returned.
• Possible settings for method include:
•  "HarmonicMean" harmonic mean of eigenvalues method "HarrisStephens" Harris–Stephens corner detection {"HarrisStephens",k} Harris–Stephens method with sensitivity parameter k "ShiTomasi" minimum eigenvalue method (default)

ExamplesExamplesopen allclose all

Basic Examples  (2)Basic Examples  (2)

Corner points in an image:

 In[1]:=
 Out[3]=

Compute and visualize the main 30 corners in an image:

 In[1]:=
 Out[1]=