IInkRecognitionResult Interface
IInkRecognitionResult Interface |
Represents the result of the recognition. The results of recognizing handwritten ink are returned in a IInkRecognitionResult object.
Members
| Enumeration | Description |
|---|---|
| InkRecognitionConfidence | Defines values that indicate the level of confidence that the recognizer has in the accuracy of the recognition result. |
| Method | Description |
|---|---|
| GetAlternatesFromSelection | Returns the alternates collection from a selection within the best result string of the recognition result so that each alternate corresponds to only one segment of ink. |
| ModifyTopAlternateMethod | Modifies the recognition result with a known alternate. |
| SetResultOnStrokes | Assigns the recognition results to the strokes that were used to create the results. |
| Property | Description |
|---|---|
| Strokes | Gets the InkStrokes that was used by the recognizer to generate the IInkRecognitionResult object. |
| TopAlternate | Gets the top alternate of the recognition result. |
| TopConfidence | Gets the confidence level of the TopAlternate property of the IInkRecognitionResult object. |
| TopString | Gets the result text for the TopAlternate property. |
Remarks
IInkRecognitionAlternate objects, or alternates, are created from the result. The best, or top, alternate is the one that is used by the default in the result. However, you can use the methods of the IInkRecognitionResult object to specify which alternates to use in the result.
System performance can suffer if recognition results are automatically assigned to every collection of strokes. Therefore, by default, results are not attached to a collection of strokes. You must call the SetResultOnStrokes method to assign results to a collection of strokes.
The only way to persist recognition results is to call SetResultOnStrokes and then add this collection of strokes to the CustomStrokes collection on the InkDisp object.
Not all recognizers set the TopConfidence property. When an application attempts to access a property that is not set by the recognizer, an argument exception is thrown.
Note: The various handwriting recognizers shipped from Microsoft® in both Latin script and East Asian languages can sometimes produce the Unicode value 0xFFFF as the recognition result. This occurs when the recognizer is unable to match a piece of ink with any valid character. The 0xFFFF code point is valid UCS-2, but not allowed in UTF-8. An application that converts recognition results to UTF-8, should replace 0xFFFF with some other code point, for example, 0xFFFD.