Chapter 7 - Tablet PC Platform SDK: Ink Recognition

Chapter 7

Tablet PC Platform SDK: Ink Recognition

One of the most important design principles followed in developing software for the Tablet PC is to treat ink as a first-class data type. This means that whenever possible, ink should remain in a document as ink rather than be converted into another data type, such as a text string. However, the ability to translate electronic ink into a text string or a command can still be of great use to a tablet-enabled application. Interoperability with other applications, the repurposing of data into other forms such as an e-mail message or a schedule entry, and searching through notes are examples of the many tasks that benefit considerably from the conversion of electronic ink into a text string.

Gesture-based commands ink strokes forming predefined shapes that trigger application behaviors also streamline the usage model of ink-enabled applications. The process of ink recognition is therefore extremely important, and not in opposition to the ink as ink philosophy; in fact, it is complementary.

Until the Tablet PC Platform, the lack of accurate handwriting recognition had plagued most, if not all, pen-based computing systems. The Tablet PC boasts the most accurate handwriting recognition ever achieved in the industry, and it supports numerous languages, including English (both U.S. and U.K.), French, German, Chinese (both simplified and traditional), Japanese, and Korean. More are on the way, too expect Spanish support to be released sometime in 2003.

This chapter discusses the various ways in which ink recognition can be performed and how the results can be interpreted. We ll start off by looking at the recognition architecture of the Tablet PC Platform and then discuss the different techniques by which ink recognition can be performed from the simplest to the most complex. Methods and tips to improve the accuracy of the recognized results maximizing the end-user experience are included throughout the chapter.