Predictive Text Fitting

摘要

A new predictive text fitting algorithm is introduced in this paper. Through this algorithm, the ideal font size of given text content can be efficiently determined to fit the text into frames of arbitrary shapes. After trying out the initial font size, the algorithm measures a fit factor based on ratio of the area used and the area needed. Then it predicts the next font size according to the fit factor, and it also adjusts the key model parameter based on previous iterations. Compared with methods that change the font size at fixed amount each time or use predetermined models, the advantages of this algorithm include fast convergence, as well as insensitivity to text placement engine and initial values of parameters. This algorithm has a number of potential applications, such as variable data printing, multimodal publishing, and adaptive graphic presentation.