Automatic Recognition of Asphalt Pavement Cracks Based on Image Processing and Machine Learning Approaches: A Comparative Study on Classifier Performance

Table 1

Prediction result comparison.

Statistical; Measure	CAR (%)	Pavement Crack Classification Models
		NBC	CT	BPANN	RBFNN	SVM	LSSVM
Average	AC	65.89	81.72	80.33	70.17	93.78	95.33
	DC	66.94	68.28	84.44	70.28	92.17	94.00
	LC	78.78	73.89	82.94	74.94	88.06	87.67
	NC	87.94	79.67	85.44	87.94	91.00	91.17
	TC	78.17	80.67	90.78	70.72	94.56	94.94
	Overall	75.54	76.84	84.79	74.81	91.91	92.62
Standard Deviation (Std.)	AC	7.18	4.09	5.94	8.07	4.17	3.46
	DC	3.48	6.38	5.88	5.03	3.31	3.02
	LC	4.75	6.21	4.08	5.28	4.78	4.48
	NC	4.24	5.24	5.57	4.37	3.65	3.56
	TC	4.53	4.96	3.27	6.60	3.36	2.08
	Overall	1.90	2.19	2.32	2.44	1.53	1.46

Note: AC: alligator crack; DC: diagonal crack; LC: longitudinal crack; NC: noncrack; TC: transverse crack.