Research Article

KmsGC: An Unsupervised Color Image Segmentation Algorithm Based on -Means Clustering and Graph Cut

Table 1

Quantitative evaluation result of algorithms on the four indices (the best values for each index are in bold letters). PRI ranges between 0, 1 , the higher the better. VoI ranges between 0, , the lower the better. GCE ranges between 0, 1 , the lower the better. BDE ranges between 0, , in the unit of pixel, the lower the better.

ApproachPRIVoIGCEBDE

Humans0.87541.10400.07974.994
KmsGC0.77122.56160.29329.1225
CTM 0.75612.46400.17679.4211
CTM 0.76272.20350.18469.4902
CTM 0.76172.02360.18779.8962
MS0.75502.47700.25989.7001