Moving Object Classification Using 3D Point Cloud in Urban Traffic Environment

Table 4

Performance comparison of moving object classification algorithms.

Object	Features	Classifier	AUC	Run time per object (ms)
Vehicle	18	SVM	0.950	40.67
		SVM-BDT	0.989	77.89
	68	SVM	0.980	60.45
		SVM-BDT	0.991	88.38
Pedestrian	18	SVM	0.953	39.77
		SVM-BDT	0.976	72.38
	68	SVM	0.969	57.42
		SVM-BDT	0.985	83.93
Bicycle	18	SVM	0.859	42.94
		SVM-BDT	0.958	75.83
	68	SVM	0.902	59.94
		SVM-BDT	0.968	85.01
Crowd	18	SVM	0.845	41.70
		SVM-BDT	0.931	75.33
	68	SVM	0.923	58.87
		SVM-BDT	0.933	84.15