Improving Readability of Contextualized Text Explanations

摘要

DUAL-USE OF IMAGE SPACE (DUIS) is an interactive technique for presenting text corresponding to images within the image space. From a technical point of view, the pixels in the image space are used both as text which can be read as well as for shading. This approach raises a number of interesting new readability problems: First, in order to simulate shading, the weight and the width of the character glyphs are manipulated. We have noted that readers find reading text with weight and width variations not only difficult but also irritating. Second, the shape of the silhouettes of the objects, by their irregular nature, are not ideal for text layout. In this paper we present techniques for addressing the DUIS readability problems. The techniques are divided into three categories as follows: First, shading and reading functionalities have been separated by creating two text presentation modes, shading mode and reading mode. In reading mode, it is no longer necessary to vary the weight and width of the character glyphs. Second, we have introduced the concept of multiple column presentation to address the problems of interruptions in reading. Finally, distortions of the objects are also used to improve readability.