Multiresolution Segmentation
Multiresolution Segmentation groups areas of similar pixel values into objects. Consequently homogeneous areas result in larger objects, heterogeneous areas in smaller ones. In Definiens's multiresolution segmentation the homogeneity criterion is used to determine which heterogeneity attributes of image objects are to be minimized as a result of a segmentation run.
Three criteria are used to describe image object heterogeneities:
- color criterion,
- smoothness criterion and
- compactness criterion.
The smoothness and the compactness criteria are additionally summarized to the shape criterion. The composition of the entire homogeneity criterion based on the specific criteria can easily be defined by assigning weights to each of the specific criteria. The multiresolution segmentation algorithm consecutively merges pixels or existing image objects. Thus it is a bottom-up segmentation algorithm based on a pairwise region merging technique. Multiresolution segmentation is an optimization procedure which, for a given number of image objects, minimizes the average heterogeneity and maximizes their respective homogeneity. Scale parameter an important parameter of the Multiresolution Segmentation algorithm, used to determine the upper limit for a permitted change of heterogeneity throughout the segmentation process. The scale parameter determines the average image object size.