Table 13.

Comparison of classification accuracies on the FLAVIA dataset with 32 species

Descriptor	Feature	Classifier	Accuracy	Study
Hu moments	Shape	SVM	25.30	[111]
HOG	Shape		84.70
SIFT	Shape		87.50	[81]
SMSD, $A_{vein}/A_{leaf}$	Shape + vein	PNN	90.31	[144]
SMSD	Shape		70.09
PFT	Shape	k-NN	76.69	[116]
SMSD, FD	Shape		84.45
SMSD, FD, CM	Color + shape	k-NN, DT	91.30
SMSD	Shape	PNN	91.40	[58]
SMSD, $A_{vein}/A_{leaf}$	Shape + vein	SVM (k-NN)	94.50 (78.00)	[5]
SIFT	Shape	SVM	95.47	[59]
SURF	Shape	SVM	95.94	[103]
SMSD, FD	Shape	BPNN	96.00	[1]
SMSD, CM, GLCM, $A_{vein}/A_{leaf}$	Shape + color + texture + vein	SVM	96.25	[48]
SMSD	Shape		87.61 (82.34, 80.26, 72.89)	[16]
SMSD, CM	Shape + color	RF (k-NN, NB, SVM)	93.95 (92.46, 88.77, 86.50)
SMSD, CM, CH	Shape + color		96.30 (94.21, 89.25, 92.89)
SMSD	Shape	NFC	97.50	[24]
CT, Hu moments	Shape		50.16 (41.60)	[23]
GF, GLCM	Texture	NFC (MLP)	81.60 (87.10)
CT, Hu moments, GF, GLCM	Shape + texture		97.60 (85.60)
EnS and CDS	Shape + texture	SVM	97.80	[140]