A new morphological approach for reducing speckle noise in ultrasonic images

Abstract

Ultrasonic images are often contaminated with speckle noise. The speckle noise makes it difficult to identify features of interest that are typically small in nature. Morphological operations have a demonstrated track record of success in minimizing speckle noise in gray-scale images. Prior attempts to solve the problem have focused on the use of multi-resolution morphological filters with a variety of structuring element sizes. Such algorithms tend to be overly complex and computationally expensive to implement. This paper proposes an efficient approach for reducing speckle noise without eliminating small sized features from the image. The basic noise reduction algorithm is implemented using a sequence of simple morphological closing operations and estimating proper threshold levels. The validity of the approach is demonstrated using several ultrasonic nondestructive evaluation images.