Computer Vision Tricks with OpenCV and Other Computer Vision Libraries

Projects

• qrcode_science: Detect QR codes with OpenCV erosion and dilations
• polygon_science: Mask regions in a certain HSV color range and calculate polygons with OpenCV and Shapely

Color Space

To find a color in HSV see the colorspace plot below. In OpenCV hue (H) is in the range 0-179, saturation (S) 0-255 and value (V) 0-255. In the plot below, V is set to a constant value of 255.

See hsv_colorspace_plotter.py to see how this image was made with the matplotlib library.

Setup

Create a Python environment (conda, venv, uv, etc.) and install the libraries with pip as follows.

pip install -r requirements.txt

Use scripts in the utils folder to help you choose color ranges.

Helper scripts in utils

To explore interactively upper and lower values for the HSV values (the hue, saturation and value), the app/tool interactive_color_thresholder.py, in the utils folder, for any image Open CV can read.

To determine upper and lower values for colors using RGB values, the find_hsv_ranges.py will provide this (check the script to set input RGB values).

To explore HSV masks and image results try out explore_hsv_masks.py (check the script to set the color ranges).