An supervised learning method for overlapping cells

DOI

10.2991/icmra-15.2015.207How to use a DOI?

Keywords

overlapping; non-rigid registration; over-segmentation; template matching

Abstract

The clustering phenomenon often appears in histopathology image, some cells overlap or touch together to from a big area. It is necessary to design an effective algorithm to separate the clustering cells into single one. We describe a generic method for segmentation microscopy images based on supervised modeling. The main idea is to use the example input segmentations to learn a statistical model of the shape and texture of the structures to be segmented. The segmentation of the test image can be functioned by maximizing the normalized cross correlation between the model and neighborhoods in the test image, accompanied by a final adjustment that utilizes nonrigid registration. This method can effectively and efficiently solve the overlapping and over-segmentation problem.

Copyright

© 2015, the Authors. Published by Atlantis Press.

Open Access

This is an open access article distributed under the CC BY-NC license (http://creativecommons.org/licenses/by-nc/4.0/).

Download article (PDF)

TY  - CONF
AU  - Pengfei Shen
AU  - Jie Yang
PY  - 2015/04
DA  - 2015/04
TI  - An supervised learning method for overlapping cells
BT  - Proceedings of the 3rd International Conference on Mechatronics, Robotics and Automation
PB  - Atlantis Press
SP  - 1071
EP  - 1075
SN  - 2352-538X
UR  - https://doi.org/10.2991/icmra-15.2015.207
DO  - 10.2991/icmra-15.2015.207
ID  - Shen2015/04
ER  -