Moving Vehicle Detection and Classification Using Gaussian Mixture Model and Ensemble Deep Learning Technique

Table 5

Performance analysis of the proposed ensemble deep learning technique on MIO-TCD in terms of FDR and FOR.

Feature extraction	Classifier	FDR (%)	FOR (%)
SPT	MSVM	22	18.03
	KNN	12	17
	DNN	9.72	10.74
	LSTM	8	3.20
	Ensemble	6	1.29
WLD	MSVM	11.91	9.5
	KNN	6.03	5.93
	DNN	3	3.07
	LSTM	2.09	1.02
	Ensemble	1.72	0.86
Hybrid (SPT + WLD)	MSVM	4	3.29
	KNN	1.20	1.95
	DNN	0.98	0.83
	LSTM	0.79	0.35
	Ensemble	0.44	0.32