Table 2.

Comparison of the number of training, iteration, time complexity, Accuracy, and time of testing with the traditional methods

Algorithm	No. of training images ($N_{\mathit{train}})$	No of iteration ($N_{\mathit{iter}})$	Time complexity ($T_{\mathit{train}})$ (s)	Accuracy (%)	Time of testing ($T_{\mathit{test}})$ (s)
CNN	15	9	4.725	73.1	7.81
	20	12	6.3	76.4
	25	16	8.4	79.12
	30	20	10.5	83.56
	35	24	12.6	86.72
	40	27	14.175	88.5
R-CNN	15	7	2.1875	74.32	5.42
	20	9	2.8125	78.64
	25	12	3.75	81.45
	30	15	4.6875	85.92
	35	18	5.625	88.62
	40	22	6.875	90.32
Fast R-CNN	15	6	1.86	78.12	3.17
	20	8	2.48	81.23
	25	10	3.1	84.32
	30	13	4.03	87.52
	35	16	4.96	89.32
	40	20	6.2	91.24
Proposed Faster R-CNN	15	4	1.1648	84.45	2.64
	20	6	1.7472	86.32
	25	8	2.3296	88.54
	30	11	3.2032	91.74
	35	13	3.7856	93.92
	40	16	4.6592	93.98