• A set of programs that are widely used in optical/microwave Remote Sensing
• Image processing tools of RGB extractor/merger, contrast enhancer, Gaussian filter, Sobel filter, etc.
• NDVI feature generator for vegetation detection
• K-Means unsupervised classification program

*Third-party Python Libraries used: matplotlib, numpy, opencv-python, Pillow, scikit-image, scikit-learn, scipy and tqdm To install the package:

python -m venv venv
source venv/bin/activate
pip install .

• ROI Clips the Region of Interest (ROI) defined by (left, right, top, bottom) pixel coordinates from an input image. The input path may be either a directory or a file. The flip argument is used to flip the image left to right. If the coordinates are not pre-defined in the commandline, press c in the preview window and define the coordinates interactively. Erase the undesired parts in the image using the e key. Press q in the preview window if the clipped image is satisfied.

  ROI --input [input path] --left [left coordinate] --right [right coordinate] --top [top coordinate] --bottom [bottom coordinate] --flip {True/False} --extension {jpg, png, tiff}
  

• RGB_extract Extracts RGB components from a True Colour Image (TCI).

  RGB_extract --input [input path] --red [red band name] --blue [blue band name] --green [green band name] --extension {jpg, png, tiff}
  

• RGB_merge Merges three grayscale images into a single TCI image.

  RGB_merge --input [input path] --red [red band (ex. B04)] --blue [blue band (ex. B03)] --green [green band (ex. B02)] --extension {jpg, png, tiff}
  

• downscale Downscales the image dimension. The divisor parameter determines the level of downscaling

  downscale --input [input path] --divisor [integer] --extension {jpg, png, tiff}
  

• contrast Enhances the contrast of a grayscale image.

  contrast --input [input path] --extension {jpg, png, tiff}
  

• gaussian Low-pass Guassian filter.

  gaussian --input [input path] --extension {jpg, png, tiff}
  

• SOBEL High-pass Sobel filter.

  SOBEL --input [input path] --extension {jpg, png, tiff}
  

• NDVI Calclates and creates a grayscale Normalized Difference Vegetation Index (NDVI) image using optical NIR and VIS band images. NDVI = (NIR — VIS)/(NIR + VIS) (Weier and Herring, 2000).

  NDVI --input [input directory path] --NIR [NIR band] --VIS [VIS band] --extension {jpg, png, tiff}
  

• K_Means Unsupervised K-Means classification. The module utilizes the scikit-learn's KMeans function (Pedregosa et al. 2011)

  K_Means --input [input directory path] --features [spaced delimited list of feature names (ex. NIR VIS)] --num_classes [# classes] --extension {jpt, png, tiff}
  

  

• GLCM Generates GLCM products (entropy, energy, contrast, homogeneity and dissimilarity) following the methods suggested by Ressel et al. (2015).

  GLCM --input [input image path] --products [spaced delimited list of desired products] --window_size [GLCM window size] --extension {jpt, png, tiff}
  

Pedregosa, F., Varoquaux, G., Gramfort, A., Michel, V., Thirion, B., Grisel, O., Blondel, M., Prettenhofer, P., Weiss, R., Dubourg, V., Vanderplas, J., Passos, D., Brucher, M., Perrot, M., & Duchesnay, E. (2011). Scikit-learn: Machine Learning in Python. Journal of Machine Learning Research, 12, 2825–2830.

Ressel, R., Frost, A., & Lehner, S. (2015). A neural network-based classification for sea ice types on X-band SAR images. IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing, 8(7), 3672-3680.

Weier, J., & Herring, D. (2000). Measuring vegetation (ndvi & evi). NASA Earth Observatory, 20.