Skip to content

Latest commit

 

History

History
88 lines (63 loc) · 4.54 KB

README.md

File metadata and controls

88 lines (63 loc) · 4.54 KB
Raw

SHINE_color

See release notes below. Please, send suggestions and doubts to [email protected]

SHINE_color was adapted from the SHINE toolbox and allows the control of low-level properties of colorful images. It does so by either manipulating RGB channels directly or by converting RGB into HSV or CIELab color space, extracting the luminance channel, applying SHINE controls, and concatenating it with the other channels (i.e., Hue, Saturation) to create a colorful image with controlled luminance.

SHINE documentation (see a manual here) extends to SHINE_color. See a step-by-step on how to use SHINE_color following.

STEP-BY-STEP

If you have no experience with MATLAB, just follow these steps (images available on the files tab of the OSF project):

  1. Download/clone the SHINE_color & unzip it on the desired folder;
  2. Go into the SHINE_color/toolbox subfolder;
  3. Add the images/videos to be processed in the "SHINE_color_INPUT" folder;
  4. Open MATLAB and select the "SHINE_color" folder, then the "SHINE_color/toolbox" subfolder;
  5. Type "SHINE_color" (case sensitive);
  6. Follow the prompts and select the operations you would like;
  7. Once it is done (the sign ">>" is back on the editor), check the "SHINE_color_OUTPUT" folder. There you will find your processed images/videos and some statistics. Also check the input folder for pre-processing statistics.

Please note that SHINE_color does not read transparent (alpha) channels from .PNG images. If you want to display images with transparent background on your experiment, upload them to SHINE_color, perform manipulations on background and foreground separately, then remove the background on an image manipulation software (e.g., GIMP, Photoshop).

References
Dal Ben, R. (2023). SHINE_color: controlling low-level properties of colorful images. MethodX, 11, 102377. https://doi.org/10.1016/j.mex.2023.102377

Willenbockel, V., Sadr, J., Fiset, D., Horne, G. O., Gosselin, F., & Tanaka, J. W. (2010). Controlling low-level image properties: The SHINE toolbox. Behavior Research Methods, 42(3), 671–684. http://doi.org/10.3758/BRM.42.3.671
SHINE toolbox is available at: http://www.mapageweb.umontreal.ca/gosselif/SHINE/

Update, April 2023, version 0.0.5

Updates & improvements:

  • Functional command line call, input is read by readImages;
  • Streamline readImages.m
  • Streamline lum2scale.m;
  • Remove image reading and preprocessing from individual functions;
  • Streamline comments, descriptions, and standardize function naming;
  • Add license info to main script;
  • Add Command Window log (diary);
  • Add message redirecting users to SHINE in case of greyscale input;
  • Make main script modular, added: -- displayInfo.m; -- processImage.m; -- userWizard.m;
  • Add RGB colorspace: -- RGB added as a cs option (SHINE_color); -- Transformations applied to each RGB channel; -- diagPlots on each RGB channel; -- lumCalc on each RGB channel; -- Provide RMSE and SSIM to each RGB channel;

Update, October 2021, version 0.0.4

Updates & improvements:

  • lum_calc is calculated directly from the input and output luminance channel. Previous versions re-read rgb images, transformed it to hsv or CIELab, and then calculated statistics. The new function is more accurate and faster;
  • diag_plots plots luminance information directly from the input and output luminance channel. The previous versions re-read rgb images, transformed it to hsv or CIELab, and then plotted the luminance information. The new function is more accurate and faster.

Update, September 2021, version 0.0.3

Updates & improvements:

  • Require input to every prompt (except for prompts with default values);
  • When dealing with images, require at least 2 images to advance;
  • Fix the pooled SD calculation from lum_calc;
  • Update lum_calc output, now with pre vs. pos summary in a single file;
  • Add option for CIELab colorspace;
  • Update functions' input to account for new colorspace (e.g., sfPlot, spectrumPlot);
  • v2scale is now lum2scale;
  • scale2v is now scale2lum;
  • Add DIAGNOSTICS subfolder in SHINE_color_OUTPUT, for storing img stats and diag plots;
  • Add a new function diag_plots for diagnostic plots of operations with images.

Update, April 2019, version 0.0.2

The new version of the SHINE_color now handles video files. If a video file is provided, all frames will be extracted, SHINE_color operations will be performed on each frame, and the video will be re-created with the manipulated frames.