-
Notifications
You must be signed in to change notification settings - Fork 42
MORF and Random Gabor Filters
https://github.com/neurodata/ProgLearn/issues/64
Random Gabor filters are a type of convolutional kernel that can extract information about textures from images. The goal of the issue was to add functionality to MORF in order to leverage random gabor filters instead of random patches. This has been put on hold, as the current SPORF repo is not sklearn compliant. Merging SPORF and MORF into sklearn has been prioritized, and this issue will be tackled afterwards.
C++ Code has been created to sample random gabor filters as done in OpenCV. Code and results are available in this repository.
C++ Code has been created to randomly sample parameters for the random gabor filters. This code can be found in this file.
C++ Code has been created to generate a bank of random gabor filters. This code can be found in this file.
The code to create random gabor filters is functional and has been proven to output the same results as using OpenCV. The other two files have not been tested enough to prove correctness. These two were written based on the previous implementation in R which can be found in this file.