Disclosed is a processor-implemented method for processing images. The processor receives image data of a color space defined by input provided to a system by a user, determines at least one color attribute of the pixels in the received image correlating to at least perceptual image characteristics, determines statistics using the attribute(s), and analyzes the statistics to classify the image into a category. Based on at least the category, a billing structure for the image is determined. For example, chroma and/or hue of pixels can be used to create histograms, whose data is used to determine a degree of color and/or content of an image, which is categorized. Color space components of received pixels can also be statistically analyzed. Such determinations consider billing based on human perception of use of color. Billing for color images in this manner satisfies the user and increases use of color output (e.g., printing).