Remote-Sensing Image Tile-Level Annotation Based on Discriminative Features and Expressive Visual Word Descriptors

Table 4

The OA and Kappa comparison between using four different methods on the first dataset.

Methods	OA	Kappa
	0.820 2	0.755 1
OB-LDA	0.874 6	0.818 8
LEGION + SVM	0.939 0	0.834 5
Our proposed method	0.942 5	0.839 3