Skip to main content

Table 1 Comparison of various denoising methods by PSNR (dB)

From: Magnitude-phase of the dual-tree quaternionic wavelet transform for multispectral satellite image denoising

Methods

Noise variance

 

15

20

25

30

40

50

Image 1

 

 Noisy image

24.87

22.40

20.78

18.19

16.69

14.91

 VisuShrink

22.93

21.51

21.49

19.71

19.90

17.90

 NeighShrink

25.23

23.43

22.89

21.93

20.63

19.58

 BiShrink

25.32

24.48

23.19

23.58

22.43

21.93

 Proposed algorithm

26.24

25.65

25.02

24.36

23.12

22.99

Image 2

 

 Noisy image

24.90

22.42

20.76

18.18

16.64

14.82

 VisuShrink

24.28

21.72

21.68

20.82

20.01

19.65

 NeighShrink

27.37

24.95

24.02

22.31

19.80

23.12

 BiShrink

26.74

25.47

25.01

20.48

23.40

23.53

 Proposed algorithm

27.74

25.34

24.90

23.95

23.36

22.63

Image 3

 

 Noisy image

25.07

22.58

20.90

18.23

16.63

14.81

 VisuShrink

24.25

22.29

25.04

22.92

22.65

21.87

 NeighShrink

27.32

26.26

25.45

22.59

22.46

19.13

 BiShrink

28.22

27.37

25.15

23.29

23.26

23.21

 Proposed algorithm

28.02

28.42

26.55

25.02

25.03

24.92

  1. Wavelet transform: dual-tree quaternionic wavelet transform with five levels.