fastGLCM

'GLCM' Texture Features

Two 'Gray Level Co-occurrence Matrix' ('GLCM') implementations are included: The first is a fast 'GLCM' feature texture computation based on 'Python' 'Numpy' arrays ('Github' Repository, <https://github.com/tzm030329/GLCM>). The second is a fast 'GLCM' 'RcppArmadillo' implementation which is parallelized (using 'OpenMP') with the option to return all 'GLCM' features at once. For more information, see "Artifact-Free Thin Cloud Removal Using Gans" by Toizumi Takahiro, Zini Simone, Sagi Kazutoshi, Kaneko Eiji, Tsukada Masato, Schettini Raimondo (2019), IEEE International Conference on Image Processing (ICIP), pp. 3596-3600, <doi:10.1109/ICIP.2019.8803652>.

README

# GLCM
Fast Gray-Level Co-Occurrence Matrix (GLCM) by numpy. This script calculates GLCM without per pixel For loop, and works faster than GLCM on scikit-image.

![img](./img/output.jpg)

## Usage

Clone or download fast_glcm.py. Use scripts as follows.

```python
import fast_glcm
from skimage import data

if __name__ == '__main__':
    img = data.camera()
    glcm_mean = fast_glcm.fast_glcm_mean(img)
```

Versions across snapshots

Dependencies (latest)