EDGE DETECTION OF DIGITISED HISTOPATHOLOGICAL SLIDE IMAGES USING DYNAMIC THRESHOLDING
Sanjay Nag, Roshni Dasgupta, Sayan Dutta, Dr. Indra Kanta Maitra, *Prof. Samir Kumar Bandyopadhyay
Malaria is one of the most common parasite transmitted disease of humans in the modern world. CAD (Computer Aided Design) which is relatively a young technology has been deemed highly impactful in recent times because of its cost effectiveness and accuracy. For the development of CAD to detect abnormalities in histopathological medical images, effective image segmentation of the image is required. For the image segmentation to be accurate it must be preceded by a robust edge detecting algorithm. In this paper we have introduced a newtree based approach, using statistical analysis and distance thresholding to provide a more efficient algorithm towards edge detection. The proposededge detection algorithm has provided a comparableresultwith other well-accepted methods.
Keywords: Histogram, BMP, full and complete binary tree, Bin variance Calculation(BVC), Maximum distance Thresholding(MDT), Prominent Bins, Truncated Bins.
[Full Text Article]