Segmentation of Microscopic Nuclei Image Using U-Net Deep Networks

Authors

  • D. Saraswathi , R. Preethi , B. Lakshmi Priya , S. Vijayananth

Abstract

Image segmentation plays a vital role in analyzing an image particularly in medical field and it is helping the
researchers to extract more information from the segmented image. Segmentation of nuclei images plays a critical
role in medical diagnosis. However, they often preferred UNet: Convolutional neural networks for segmentation.
The train and test images are given to UNet for analyzing the segmentation performance. The UNet architecture
consists of contraction and expansion path for precise segmentation. This method outperforms the existing
segmentation techniques in segmenting the nuclei structure in electron microscope. Comparatively computational
complexity is less.

Published

2020-03-31

Issue

Section

Articles