Table 5.

Results of different segmentation methods on commonly used crack datasets.

Method	CrackLS315		DeepCrack537		CFD
	MIoU	F1-Score	MIoU	F1-Score	MIoU	F1-Score
U-Net	65.30	62.54	83.70	81.52	66.79	72.44
DeepLabv3+	63.52	67.59	82.84	79.82	74.75	83.13
SETR	70.51	70.08	81.20	81.34	63.26	69.74
SegFormer	62.29	68.99	85.52	89.31	75.98	83.87
Mask2Former	74.65	79.47	85.92	87.67	77.86	82.25
CrackFormer-II	89.50	88.68	87.60	86.97	83.98	87.42
SAM	69.20	78.05	84.59	83.86	72.44	76.40
CrackCTFuse	89.65	85.40	87.03	89.43	84.82	90.59