Abstract

A switching median filter is effective for impulse noise elimination while preserving edges and details of an image. In the switching median filter an impulse noise detector is employed before filtering, and the detection result is used to control whether a pixel should be filtered or not. However, the conventional impulse detector tends to misjudge noise-free pixels constructing line structures to be the noises. We propose a new random-valued impulse noise detector based on the minimum spanning tree, and it is applied to the switching median filtering to eliminate the impulse noise effectively even for the image including line structures. Through the experiments, the effectiveness of the proposed random-valued impulse noise detector is illustrated.

© 2008 Optical Society of America

Full Article  |  PDF Article

References

  • View by:
  • |
  • |
  • |

  1. T. Sun and Y. Neuvo, Pattern Recogn. Lett. 15, 341 (1994).
    [CrossRef]
  2. Z. Wang and D. Zhang, IEEE Trans. Circuits Syst., II 46, 78 (1999).
    [CrossRef]
  3. R. J. Wilson, Introduction to Graph Theory (Pearson Education Limited, 1996).

1999 (1)

Z. Wang and D. Zhang, IEEE Trans. Circuits Syst., II 46, 78 (1999).
[CrossRef]

1994 (1)

T. Sun and Y. Neuvo, Pattern Recogn. Lett. 15, 341 (1994).
[CrossRef]

Neuvo, Y.

T. Sun and Y. Neuvo, Pattern Recogn. Lett. 15, 341 (1994).
[CrossRef]

Sun, T.

T. Sun and Y. Neuvo, Pattern Recogn. Lett. 15, 341 (1994).
[CrossRef]

Wang, Z.

Z. Wang and D. Zhang, IEEE Trans. Circuits Syst., II 46, 78 (1999).
[CrossRef]

Wilson, R. J.

R. J. Wilson, Introduction to Graph Theory (Pearson Education Limited, 1996).

Zhang, D.

Z. Wang and D. Zhang, IEEE Trans. Circuits Syst., II 46, 78 (1999).
[CrossRef]

IEEE Trans. Circuits Syst., II (1)

Z. Wang and D. Zhang, IEEE Trans. Circuits Syst., II 46, 78 (1999).
[CrossRef]

Pattern Recogn. Lett. (1)

T. Sun and Y. Neuvo, Pattern Recogn. Lett. 15, 341 (1994).
[CrossRef]

Other (1)

R. J. Wilson, Introduction to Graph Theory (Pearson Education Limited, 1996).

Cited By

OSA participates in CrossRef's Cited-By Linking service. Citing articles from OSA journals and other participating publishers are listed here.

Alert me when this article is cited.


Figures (4)

Fig. 1
Fig. 1

MST of a graph obtained from an image.

Fig. 2
Fig. 2

Schematic of the procedures of the proposed impulse noise detection. (a) Noise probability calculation, (b) noise map generation.

Fig. 3
Fig. 3

Filtering results for a synthetic test image. (a) Original image, (b) input image I R with p = 0.01 , (c) MF ( ρ = 3 ) , (d) SMF ( ρ = 3 , ϴ = 60 ), (e) SMF ( ρ = 3 , ϴ = 120 ), (f) SMF-MST ( r = 3 , θ = 0.7 , ρ = 3 ).

Fig. 4
Fig. 4

Image Parrots employed and its filtering results. (a) Original image, (b) partial image with line structures, (c) input image I R with p = 0.01 , (d) MF ( ρ = 3 ) , (e) SMF ( ρ = 3 , ϴ = 100 ), (f) SMF-MST ( r = 3 , θ = 0.9 , ρ = 3 ).

Tables (1)

Tables Icon

Table 1 MSEs of Filtering Results Concerning Parrots Corrupted by Random-Valued Impulse Noise

Equations (3)

Equations on this page are rendered with MathJax. Learn more.

I ( x , y ) = { I M ( x , y ) I ( x , y ) I M ( x , y ) > ϴ I ( x , y ) otherwise } .
I ( x , y ) = { I M ( x , y ) C ( x , y ) = 1 I ( x , y ) otherwise } .
I R ( x , y ) = { I ( x , y ) : probability 1 p R ( x , y ) : probability p } ,

Metrics