International Conference on Advances in Pattern Recognition · 2000 · 544 citations · 19 references
ComputingComputer ScienceComputer VisionDigital ImagingStatistical ComputingCluster DetectionImage AnalysisImage SegmentationCluster AnalysisCluster Analysis For SegmentationCluster Evaluation MetricsCluster Validity IndicesClusteringColor SegmentationDistance MeasuresK-means Clustering
K‑means requires the number of clusters to be specified in advance, which limits its applicability. The study proposes a simple validity measure using intra‑ and inter‑cluster distances to automatically determine the optimal number of clusters. The method generates segmentations for 2 to Kmax clusters, computes the validity measure for each, selects the clustering with the minimum value, and is evaluated on synthetic and natural images.
The main disadvantage of the k-means algorithm is that the number of clusters, K, must be supplied as a parameter. In this paper we present a simple validity measure based on the intra-cluster and inter-cluster distance measures which allows the number of clusters to be determined automatically. The basic procedure involves producing all the segmented images for 2 clusters up to Kmax clusters, where Kmax represents an upper limit on the number of clusters. Then our validity measure is calculated to determine which is the best clustering by finding the minimum value for our measure. The validity measure is tested for synthetic images for which the number of clusters in known, and is also implemented for natural images.
19
Algorithms for Clustering Data
Warren S. Sarle, Anil K. Jain, Richard C. Dubes · Technometrics · 1990 · 7.8K citations