Ultrasonic NDE images are often contaminated with speckle noise. The degradation caused by the presence of speckle noise makes it difficult to identify features of interest that are typically thin or small in nature. A variety of techniques have been proposed to date for reducing such noise. As an example, lowpass filters can be employed to reduce speckle noise. However, they tend to blur thin features and edges. Median filters are also used widely to remove impulse type noise while preserving edges in images [1]. Unfortunately, such filters perform poorly when the spatial density of the noise is high [3,6]. As an alternative, gray-scale morphological approaches involving such operations as opening, closing or combinations thereof can be applied to reduce noise in gray-scale images [1–5]. Even in this case, features that are thin or small tend to be filtered out along with the noise [6]. Prior attempts to remedy the problem have relied on the use of multi-resolution (or multi-scale) morphological filters using an array of structuring element sizes. Such algorithms tend to be overly complex and computationally expensive to implement [6].
[1]
Jean Serra,et al.
Image Analysis and Mathematical Morphology
,
1983
.
[2]
Sunanda Mitra,et al.
Optimum morphological filtering to remove speckle noise from SAR images
,
1993,
Optics & Photonics.
[3]
Petros Maragos,et al.
Morphological filters-Part II: Their relations to median, order-statistic, and stack filters
,
1987,
IEEE Trans. Acoust. Speech Signal Process..
[4]
Richard Alan Peters,et al.
A new algorithm for image noise reduction using mathematical morphology
,
1995,
IEEE Trans. Image Process..
[5]
Xinhua Zhuang,et al.
Morphological structuring element decomposition
,
1986
.
[6]
Petros Maragos,et al.
Morphological filters-Part I: Their set-theoretic analysis and relations to linear shift-invariant filters
,
1987,
IEEE Trans. Acoust. Speech Signal Process..
[7]
Edward R. Dougherty,et al.
Morphological methods in image and signal processing
,
1988
.