Alternating sequential operators from center-surround top-hat transform

Constructing alternating sequential operators is the important way of improving the efficiency of morphological operations for image processing. And, the alternating sequential filters (ASFs), which are constructed by using the morphological opening and closing, have been the important alternating sequential operators of mathematical morphology. However, because the morphological opening and closing operations have the defect of image detail smoothing, ASFs could not perform very well for some detail preserving image processing applications. Center-surround top-hat transform has been proposed through reconstructing the used structuring elements following the properties of image regions, which ensures that the main operations in center-surround top-hat transform could protect some image details; and, the main operations in center-surround top-hat transform achieve the similar function as opening and closing for bright and black feature extraction. Therefore, based on center-surround top-hat transform, some new and effective alternating sequential operators may be constructed. In light of this, a new class of alternating sequential operators with improved performance based on the center-surround top-hat transform is proposed in this paper. The definition and properties of the new alternating sequential operators are given and analyzed. Also, an application of impulsive noise suppression is used to show the improved performance of the new alternating sequential operators comparing with ASFs. Moreover, because the new alternating sequential operators constructed in this paper have some superiorities over ASFs, they could be also widely used in different applications in which the ASFs perform well, which indicates that the new alternating sequential operators may be also a class of useful morphological operators for image analysis.

► Constructing effective alternating sequential operators from center-surround top-hat transform.
► Analyzing the properties of the new alternating sequential operators.
► Performing better than ASFs in some cases.