Test out other highlighting colors

[iRove108]

Main.py for the highlight extraction example uses the default lower and upper HSV bounds, as specified in highlighting.py. These bounds have been fine-tuned for a standard yellow highlighter. This is why the example works well on this image. However, it may be possible to get the code to work for different colors as well. After installing the package as described in the README:

1. Find an image of text highlighted in a color other than the standard yellow (e.g. green). If you own highlighters, do some highlighting and take a picture.
   • Place that image in the highlighted_text directory, naming it something like image_green_highlight.jpg
2. Duplicate main.py and call it something like main_green_highlight.py
3. Edit the script to open your new image instead of image.jpg (relevant line)
4. Use the a color picker like this one to find the lower/upper bounds you would like to set to extract your new highlight color.
   • Set your new lowerb and upperb arguments when creating the HighlightExtractor (arrays of form [H, S, V])
5. Test it out! See if you can fine tune your original lower/upper bound arrays to improve the extraction

It's okay if this doesn't work perfectly! It might not—I've only tested the code with yellow highlighting colors. Nonetheless, you can still create the PR to demonstrate how well the library does/doesn't work for other colors.

[saidaibrg]

Hi Alex, I like your idea for the library! I tried testing out your code for green highlights. I was a bit confused about how to test and run the code, so here're some suggestions about this issue to help future contributors:

• Many people don't have opencv installed, maybe it's worth mentioning that they'll need to import cv2 before running the program?
• For those unfamiliar with HSV colors, links to resources about it would be helpful.
• Could you please clarify the fifth step? In my case, when I tried running python main.pyin the Terminal, I got the error saying that there'sNo module named markups

[iRove108]

The library should install OpenCV when it is installed with pip. Additionally, for others having the same problem as the previous poster mentioned in their third bullet, please see this issue.