As shown and well described here, we noticed that an illusion is produced by consistently showing the viewer a chromatic-inverted image and suddenly flashing the image in greyscale. This gives the viewers an illusion of seeing the real image, which was never shown to them. The code adds textures to the image so that the illusion is more effective and a loading bar that self-calibrates each image's dimensions. To use the illusion, stare at the black dot until the progress bar finishes - at which point the image should give the perception of being colored.
To fully automate this process, we present a project made using PILLOW. It can be accessed in the file main.py to run on your system or color_inversion_optical_illusion.ipynb which provides an interactive method to upload files and receive the illusion.
Click the file below to view the gif
from PIL import Image, ImageOps, ImageDraw
Invert the colors of the image
Color inversion, as the name suggests, is the process of inverting the RGB values of an image, an array of pixels.
If:
1. Value of initial Red = X
2. Value of initial Blue = Y
3. Value of initial Green = Z
Then:
1. Value of final Red = 255-X
2. Value of final Blue = 255-Y
3. Value of final Green = 255-Z
inverted = ImageOps.invert(img)
Convert the image to greyscale
imgGray = img.convert('L')
Add the loading bar and concatenate inverted the greyscale images!