Skip to main content

Table 7 Results of applying image filters with clustering size of 4 on SVM model

From: Automatic classification of spread‐F types in ionogram images using support vector machine and convolutional neural network

Applied clusters

Model accuracy of each cluster(s) with image filtering (%)

Image sharpening

Image thresholding

Median blur

Gamma correction

Fast-NLmeans

Bilateral

1st cluster

86.2

85.7

86.1

85.1

85.8

86.1

2nd cluster

85.7

85.8

85.5

86.1

85.9

85.8

3rd cluster

86.8

85.9

84.7

86.4

86.4

86.4

4th cluster

85.9

85.4

86.8

86.1

85.2

85.8

1st and 2nd cluster

84.9

85.7

85.1

85.7

86.4

86.2

1stand 3rd cluster

86.8

86.1

84.5

85.4

86.4

86.5

1st and 4th cluster

86.2

85.7

85.9

86.1

86.5

85.5

1st, 2nd and 3rd cluster

86.2

85.9

83.4

84.8

86.1

86.5

1st, 2nd and 4th cluster

85.4

85.8

86.1

85.8

86.2

86.5

1st, 3rd and 4th cluster

86.5

86.2

84.4

86.4

86.5

85.8

2nd and 3rd cluster

85.5

86.2

84.5

85.9

85.8

86.5

2nd and 4th cluster

85.9

85.4

85.5

85.8

86.1

86.2

2nd, 3rd and 4th cluster

86.4

85.8

86.1

85.9

85.9

86.1

3rd and 4th cluster

86.2

85.9

85.5

86.4

85.8

85.8

All clusters

86.1

86.4

83.9

85.8

85.8

85.8