Nissl-stained sections segmentation using algorithms of Mathematical Morphology

 

Fig.3 (Above) Example of segmentation results overlapped to the initial image

Fig. 3 shows segmented image at higher magnification. If you need original images to check the result, or the result of the segmentation, just say so. I always can give them to you using some file-sharing site. And please, do not hesitate to ask if you have any questions.

The code is included below. The only “unusual” dependency is MAMBA-Image (http://www.mamba-image.org/). Other libraries you probably have already. Unfortunately, Word Press does not allow uploading Jupyter Notebook  files (“ipynb”) for “security reason”. Still, you can get it here (Human_segmentation_counting_demo) as html, cut and paste it to a new file on your desktop, save as Python file and run. Also, slightly modified version of the same approach is described in my more recent article “Mouse brain neurons segmentation and counting”

Copyright © 2017-2018, Brain Imaging Informatics, LLC

One thought on “Nissl-stained sections segmentation using algorithms of Mathematical Morphology”

Leave a Reply

This site uses Akismet to reduce spam. Learn how your comment data is processed.