Measuring Visual Information

What does it mean to sequence a series of images by visual complexity?

To effectively execute this project, I employed MATLAB’s entropy function for grayscale images, one established method for quantifying an image’s visual complexity. This function assesses the degree of randomness in an image histogram, which charts the distribution of pixel intensities within that image. A high entropy value, indicating a high degree of randomness, suggests rich detail, while a low entropy value indicates simplicity due to less randomness.

Furthermore, to assess the complexity of both the structural and textural features, I exported each image at two resolutions: 500 x 400 pixels and 2,000 x 1,600 pixels. I then averaged these two results for a more comprehensive measure of entropy and used this average to sequence the images from the highest to the lowest amount of visual complexity.