Image border color extraction

Find the hex color value of the average color around an image's border

Gets the pixels from the outer 5% of an image, discards outlier values, and returns the average color in hex (as a string)

Usage

python image-decoder.py -i path/to/image.jpg

• -i input image
• -q quiet mode
• -p OFF disables the gui preview of the result

Implementation

• Downscale the source image to have a width of 200px, keeping the aspect ratio. We don't need to keep the original (large) resolution.
• Choose a percentage of the image's border on each axis to use for the color calculation (I use 5%), and put the coordinates of these pixels into a list.
• Split the scaled image to individual lists of R, G, B color channels.
• Using the coordinates list, populate 3 more lists of just the color values (0-255) for the perimeter pixels of each channel.
• Use numpy to remove statistically significant outliers from each of the lists made in the previous step.
• Calculate the average value for each channel from the result of the previous step.
• Merge the averages of each channel (an RGB(1, 2, 3) value) into hex (#123456) and return it.
• Optionally display the input image overlaid over the calculated background color using a GUI with PIL.Image.show().