PRTools contents

PRTools manual

classim

CLASSIM

Classify image and return resulting label image

    LABELS = CLASSIM(Z)
    LABELS = CLASSIM(A,W)
    LABELS = A*W*CLASSIM

Input
 Z Classified dataset, or
 A,W Dataset and classifier mapping

Output
 LABELS Label image  When no output is requested, the label image is displayed.

Description

Returns an image with the labels of the classified dataset image Z (typically, the result of a mapping or classification A*W in which A is  a set of images stored as features using IM2FEAT). For each object in  Z (a pixel), a numeric class label is returned. The colormap is loaded  automatically.

Note that if the number of classes is small, e.g. 2, an appropriate  colormap has to be loaded for displaying the result by IMAGE(LABELS),  or more appropriately, LABELS should be multiplied such that the minimum  and maximum of LABELS are well spread in the [1,64] interval of the  standard colormaps.

Example(s)

 PREX_SPATM

See also

mappings, datasets, im2feat, labeld,

PRTools contents

PRTools manual