Skip to main content

Table 3 The classification rates using the SVM classifier

From: Vehicle color classification using manifold learning methods from urban surveillance videos

Video clips

Correct classification

Number of vehicles

Accuracy rates (%)

1

191

212

90.01

2

209

238

87.82

3

196

234

83.76

4

78

91

85.71

5

501

572

87.59

6

498

552

90.22

7

779

881

88.42

8

1,355

1,547

87.59

9

1,155

1,386

83.33

10

391

447

87.47

11

94

94

1

12

956

1,150

83.13

13

835

1,019

81.94

14

749

810

92.47

15

4,551

5,226

87.08

16 (multi-lane)

7,583

8,462

89.61

17 (multi-lane)

8,266

9,278

89.09

18 (multi-lane)

9,428

10,446

90.25

Average accuracy rate

88.67