ABSTRACT
This chapter provides an overview of a number of image-processing techniques. The term image processing refers to methods that input and output an image. The chapter describes the place and role image processing has in the visualization pipeline. It presents a number of image-processing algorithms that are frequently used in data visualization. The chapter shows how higher-level representations from it, such as shapes, can be extracted and analyzed from image data. In addition to full-color images, there exists also a separate class of image formats called indexed or palette-based image formats. Image-processing operations can be applied also at earlier stages of the visualization pipeline. A general-purpose method for improving the robustness of edge detection, or for that matter of any other image-processing operation involving derivatives, is to remove small-scale noise. This corresponds to removing sharp image variations, i.e., high-frequency image components.
