Segmentation of structural parts of rosebush plants with 3D point-based deep learning methods

Table 7 Segmentation results on 8 real rosebush models from ROSE-X data set with DGCNN

DGCNN		I	II	III	S	S + I	S + II	S + III
Re	Flower	5.62	29.85	7.12	79.65	24.36	34.83	59.16
	Leaf	97.85	95.93	98.58	84.27	95.91	97.45	98.20
	Stem	9.68	28.59	27.94	49.05	46.81	73.44	67.35
Pr	Flower	22.67	54.80	75.96	42.46	37.51	75.46	79.71
	Leaf	82.34	86.80	84.60	92.26	88.05	92.71	93.27
	Stem	58.97	51.36	65.42	47.21	77.97	74.16	84.65
IoU	Flower	4.72	23.95	6.96	38.30	17.33	31.28	51.42
	Leaf	80.88	83.72	83.57	78.71	84.87	90.51	91.71
	Stem	9.07	22.50	24.34	31.68	41.35	58.48	60.02
MIoU		31.56	43.39	38.29	49.56	47.85	60.09	67.72
Acc		80.79	83.23	83.59	79.87	84.90	90.00	91.73

ISSN: 1746-4811