Computing the color complexity of images

The color characteristic of an image includes the color variety feature and the color spatial distribution feature. According to the concept of complexity in the constitution theory, two general sets are defined to describe the color variety feature of an image and the spatial distribution feature of a color respectively. Based on the definitions, an algorithm used to compute the color complexity of an image is proposed and has been implemented in the RGB color model. Several images with different contents are counted. The results show that the definition of the color complexity is effective, but the results are not consistent with the human perception due to the disadvantage of the RGB color model. Next we will work on finding a new computing method of the color complexity of images to make it consistent with the human perception.