Skip to content

Latest commit

 

History

History
45 lines (26 loc) · 3.23 KB

README.md

File metadata and controls

45 lines (26 loc) · 3.23 KB

Lumeleon v0.1.1

A Python package designed for time series analysis of color change presented in our pre-print paper "On the role of melanistic coloration on thermoregulation in the crepuscular gecko Eublepharis macularius" available here on bioRxiv.

Step 1: Use syntax "python patternAnalysis.py", to launch the user interface, where you can import folders of images to be analyzed in sequence.

Intensity Matching:

This process is intinded to standardize the lighting conditions of images taken in sequence to a color standard that is present within all images. This was created with a time series color analysis in mind so that color change can be compared across images. Note that this method does not serve to detect the specific value of color, only the relative change in color between the images.

Step 1: Use "Standardize Images", to scale the images to a color standard present in all images. The reference image will be displayed first.

Step 2: Define a reference rectangle on the image by clicking and holding the left mouse button.

Step 3: If you are happy with the crop press "y", if not, press "n" and you can repeat the process.

image

Step 4: This process will be repeated for all the images in the folder.

Result: A new subfolder called "modified" will be created containing all the modified images along with the reference image.

Segmentation

This process can potentially serve two purposes within your analysis. First, it can act to seperate the subject from the background of the image (this works best if the background is a uniform color and relatively free of excessive clutter). However, a separate manually cropping step is sufficient for this purpose. Second, the main purpose was to segment color patches into different groups and analyze color change within them.

download

Step 1: Use "Segment Images" to begin color spot segmentation via k-means clustering. It shows you the resulting masks along with the original images for visual confirmation. At this point select the number of the mask that includes the correct part of the image. In this example with 4 clusters, I select number 0. It saves the segmented image with a "_4" at the end of the name so you will be able to reproduce your work later.

download

thumbnail_test_4

Step 2: You can repeat step 1 as many times as you want to go down the hierarchical tree. For example I repeated step 2 for the above image with 3 clusters and the result was:

thumbnail_image

test_4_3

Extraction

After Segmenting, you can extract the luminance values for all images in the selected folder into a csv file.