PixelClassifier
This code is based on
https://github.com/HMS-IDAC/MLRFS
and
https://github.com/HMS-IDAC/MLRFSwCF
The main differences are:
> Only one Random Forest layer is implemented. This makes the model simpler to understand and faster to train/test.
> More feature options are available, notably steerable and log filters. This makes it useful for a wider range or problems (e.g. filament and point source detection).
> Parallel processing is implemented, both during training and segmentation. This makes it significantly faster to train/execute.
The main scripts are:
pixelClassifierTrain, used to train the model, and
pixelClassifier, used to segment images after the model is trained.
See those files for details and parameters to set.
Labels/annotations can be created with ImageAnnotationBot, available at https://www.mathworks.com/matlabcentral/fileexchange/64719-imageannotationbot
A sample dataset for a running demo is available at https://www.dropbox.com/s/hl6jvwyea9vwh50/DataForPC.zip?dl=0
This code uses 2-D steerable filters for feature detection, developed by Francois Aguet, available at http://www.francoisaguet.net/software.html
Developed by:
Marcelo Cicconet
marceloc.net
Cite As
Marcelo Cicconet (2024). PixelClassifier (https://github.com/HMS-IDAC/PixelClassifier), GitHub. Retrieved .
MATLAB Release Compatibility
Platform Compatibility
Windows macOS LinuxCategories
- Image Processing and Computer Vision > Computer Vision Toolbox > Recognition, Object Detection, and Semantic Segmentation > Object Detection Using Features >
Tags
Community Treasure Hunt
Find the treasures in MATLAB Central and discover how the community can help you!
Start Hunting!Discover Live Editor
Create scripts with code, output, and formatted text in a single executable document.
Versions that use the GitHub default branch cannot be downloaded
Version | Published | Release Notes | |
---|---|---|---|
1.0.0.0 |
|