March 3, 2020 - Seeing Replicability of Script
Lab Work
Today, I tried to see the replicability of what I have in my script so far on a new image:
Yesterday, I was able to produce a color scatter plot for this image:
But it took a very long time to produce this plot, so I was thinking of possibly finding a different way to threshold the color values in these images overall. Another issue that I came across was once I created the scatter plot, finding the colors that actually corresponded to the eggs and sperm cells seen in the image was a lot more difficult than I had hoped for it to be. Even after I had found these color value thresholds, the output was not promising:
color threshold for eggs:
I tried to run the script again using different RETR, which all produced similar results:
RETR_LIST
RETR_TREE
RETR_CCOMP
I found a source that could possibly help to threshold these color values better:
Color spaces in OpenCV (C++ / Python)
Hopefully, by improving the way I find the color threshold values, I will also be able to improve the counts of the total amount of eggs seen in the image.
Next Steps
Go through the steps seen in the source above to try and find a better way to find color threshold values for egg and sperm cells.