Tsegment - Methods to Segment an Image with Thresholding Fuzzy c-Means;/


NAME



Esegment - Methods to Segment an Image with Thresholding Fuzzy c-Means




(

SYNOPSIS



>status=SegmentImage(image,colorspace,verbose)




B

FUNCTION DESCRIPTIONS






0

SegmentImage



GMethod SegmentImage segment an image by analyzing the histograms of theJcolor components and identifying units that are homogeneous with the fuzzyc-means technique.

GSpecify Icluster threshold as the number of pixels in each cluster mustBexceed the the cluster threshold to be considered valid. SmoothingHthreshold eliminates noise in the second derivative of the histogram. AsHthe value is increased, you can expect a smoother second derivative. Thedefault is 1.5.

*The format of the SegmentImage routine is:

:

        status=SegmentImage(image,colorspace,verbose)


(A description of each parameter follows.

5
colors:


FThe SegmentImage function returns this integer value. It is the actual+number of colors allocated in the colormap.

image:


CSpecifies a pointer to an Image structure; returned from ReadImage.$

colorspace:


KAn unsigned integer value that indicates the colorspace. Empirical evidenceDsuggests that distances in YUV or YIQ correspond to perceptual colorJdifferences more closely than do distances in RGB space. The image is then1returned to RGB colorspace after color reduction.!

verbose:


JA value greater than zero prints detailed information about the identifiedclasses.