• With existing methods, tuning is required depending on the type of image.

    • By using NN, a wide range of applications can be achieved.
  • In addition to neural networks, previous studies on measuring visual complexity are also introduced.

    • 2.2. Algorithms to quantify visual complexity
      • Miniukovich and Angeli (2014) proposed five factors (visual clutter, symmetry, contour density, figure-ground contrast, and color variability) to measure visual complexity for web design and GUI applications, using spaces to evaluate the visual complexity of Chinese ink paintings.