Automatic Segmentation and Classification for Antinuclear Antibody Images Based on Deep Learning

Table 1

Accuracy obtained by different features and traditional machine learning methods.

Features	Methods
	SVM	KNN	RF	ECLF
SIFT	0.8525	0.8072	0.8417	0.8532
LBP	0.8221	0.8275	0.8640	0.8728
CoALBP	0.8539	0.8444	0.8566	0.8809
RIC-LBP	0.9039	0.8424	0.8613	0.9053
FUS (SIFT + RIC − LBP)	0.9107	0.8451	0.8978	0.9269