Pattern recognition framework for histological slide segmentation

Abstract

The venous system is similar in all people, but there are some individual variations. The system is well described and compared in the literature but it seems that there is a lack of comparison on the microscopic level. In this paper we present a segmentation framework for histological image segmentation that uses a clustering approach. The described framework can be used for further comparison of the venous system at the microscopic level. For that purpose we adopted a k–means and fuzzy c–means algorithms to classify image pixels to obtain vein segmentation. The presented results are promising and achieved partitioning can further be utilized for quantitative vein comparison.