The use of local entropy measures in edge detection for cytological image analysis

Abstract

SUMMARYAccurate edge detection is a fundamental problem in the areas of image processing and pattern recognition/classification. The lack of effective edge detection methods has slowed the application of image processing to many areas, in particular diagnostic cytology, and is a major factor in the lack of acceptance of image processing in service orientated pathology. In this paper, we present a two‐step procedure which detects edges. Since most images are corrupted by noise and often contain artefacts, the first step is to clean up the image. Our approach is to use a median filter to reduce noise and background artefacts. The second operation is to locate image pixels which are ‘information rich’ by using local statistics. This step locates the regions of the image most likely to contain edges. The application of a threshold can then pin‐point those pixels forming the edge of structures of interest. The procedure has been tested on routine cytologic specimens.

Image analysis and morphometry in the diagnosis of breast cancer

“…3a,b). We have repeatedly published on the subject (Barba et al, 1988(Barba et al, , 1989(Barba et al, , 1992Wu et al, 1996a,b,c;Wu et al, 1998a,b). In our view, the most promising procedures (Wu et al, 1998) are those based on the combination of a superimposed template (for instance, that nuclei are roughly oval and may not be configured in extravagant shapes) and multiple iterations (repeated circling along the previous boundary checking for imperfections and improving with each pass).…”

Using entropy information measures for edge detection in digital images

“…Our results strongly indicate that the better edge continuity is achievable with entropy operators then with Sobel operator. The only exception was mutual information operator (4). The better performance is due to considerable flexibility which entropy based operators have in sense of size, shape and orientation, as well as the fact that their result is a real number from predefined range.…”

“…The idea to use information entropy has been proposed more twenty five 25 years ago in digital microscopy [4], and in various forms has been in sporadic use since then [5], [6]. It is still active area of research [7], [8] wiith open possibility for improvement and wider application.…”

Segment and Fit Thresholding: A New Method for Image Analysis Applied to Microarray and Immunofluorescence Data

“…Other strategies rely on edge detection to locate signal regions, typically by finding steep intensity gradients or high spatial frequencies. 19–21 Such approaches may not be reliable where steep edges are not present in the signal regions, or where shapes are irregular. In some images, portions of true signals have sharp edges and others do not, making a single threshold in gradient or spatial frequency inaccurate in certain places.…”