A hybrid CNN and Random Forest for Spore Segmentation
Machine learning (ML) and deep learning (DL) methods can be used to analyze large amounts of TEM images in an efficient manner while also minimizing human bias and error. With ML and DL, a computer program can be trained to identify and segment the spores from TEM images using datasets of labeled images. Once the program is trained, it can be used to automatically analyze new TEM images and identify spores, which is much faster and less prone to human error than manual analysis.
- pip install tensorflow
- pip install keras
- pip install opencv-python
- pip install scikit-learn
Follow the link provided to download the trained weight file https://umeauniversity.sharepoint.com/:u:/r/sites/Spores/Shared%20Documents/General/Projects/Computer%20vision/Paper%201.%20Spore%20segmentation%20using%20CNN/VGG_300_384_1024?csf=1&web=1&e=4NTpPM
The link provided to download the training dataset https://umeauniversity.sharepoint.com/:f:/s/Spores/ErsE2RNVl8RJs4rZ2hm0y1YBLQBJa5vDgxsmUr7fzmHAyg?e=JcIdHa
Link to download test dataset https://umeauniversity.sharepoint.com/:f:/s/Spores/EucW8l4MpqRAhE5WpJLCue4BHayjRyzGHYTZPLZeMMLR-g?e=f7XaC1
License The license under which the project is released (e.g. MIT, Apache, etc.).
Acknowledge
Future Works
Conclusion