Importance-driven halftoning

Many halftoning algorithms have been proposed to simulate the original continuous tone images as closely as possible. Most of these techniques are designed to preserve particular image attributes. This thesis proposes a new generalized halftoning technique based on preserving artifacts in the image that have been predetermined by the user to be important. This new technique allows for both diversity in specifying the important attributes and control over local placement of ink. The flexibility of this technique allows for the creation of many different results. This thesis does not address what image attributes are important or how to display them, rather it introduces a new technique called importance driven halftoning. Results obtained by varying the parameters of this new technique will be illustrated and compared with traditional halftoning results, limited resource renderings and non-photorealistic rendering results.