Author:
Fend Chiara,Moghiseh Ali,Redenbach Claudia,Schladitz Katja
Abstract
AbstractReconstruction of highly porous structures from FIB-SEM image stacks is a difficult segmentation task. Supervised machine learning approaches demand large amounts of labeled data for training, that are hard to get in this case. A way to circumvent this problem is to train on simulated images. Here, we report on segmentation results derived by training a convolutional neural network solely on simulated FIB-SEM image stacks of realizations of a variety of stochastic geometry models.
Publisher
Springer Berlin Heidelberg
Reference23 articles.
1. Altendorf, H., Jeulin, D.: Random-walk-based stochastic modeling of threedimensional fiber systems. Phys. Rev. E 83, 041804 (Apr 2011)
2. Bezrukov, A., Bargie l, M., Stoyan, D.: Statistical analysis of simulated random packings of spheres. Part. Part. Systems Char. 19, 111–118 (2002)
3. Bezrukov, A., Stoyan, D., Bargie l, M.: Spatial statistics for simulated packings of spheres. Image Anal. Stereol. 20, 203–206 (2001)
4. C¸ i¸cek, ¨ O., Abdulkadir, A., Lienkamp, S.S., Brox, T., Ronneberger, O.: 3D U-Net: learning dense volumetric segmentation from sparse annotation. In: International Conference on Medical image computing and computer-assisted intervention. pp. 424–432. Springer (2016)
5. Fend, C., Moghiseh, A., Redenbach, C., Schladitz, K.: Reconstruction of highly porous structures from FIB-SEM using a deep neural network trained on synthetic images (2019), in preparation.
Cited by
2 articles.
订阅此论文施引文献
订阅此论文施引文献，注册后可以免费订阅5篇论文的施引文献，订阅后可以查看论文全部施引文献