U_Net_Ni_WC_Optical_images

Published: 4 May 2022| Version 1 | DOI: 10.17632/2wmbc95xy9.1
Contributor:
Dylan Rose

Description

This data was used to train a U-Net convolutional neural network to semantically segment WC from Ni-WC optical microscopy images. The training and test datasets contain 194 and 37 512x512 pixel images and their corresponding ground truths, respectively.

Files

Institutions

  • University of Alberta

Categories

Computer Vision, Image Segmentation, Metal Matrix Composite, Light Microscopy

Licence