Harmonious graphics generating based on the 1/f function theory

In the 1970s, Richard Voss and John Clarke researched the actual audio physical sound of music. There are three types of noise: white noise, 1/f noise, and Brownian motion noise (1/f²). 1/f noise is found to be most pleasing to human ears. White noise is too random and Brownian noise is too correlated. Similarly, for 2-dimensional sources such as graphics and images, three characteristics of monotonous, harmony and muss are also observed. The 1/f fluctuation theory provides a good way to generate affective signals both for 1-dimensional and 2-dimensional signals. This paper provides an algorithm which can generate affective patterns or graphics and obey the criteria of affective information processing. © 2006 Elsevier Ltd. All rights reserved.