This Help topic refers to the following editions:
þ Enterprise þProfessional þ Personal þ Small Business
Training the OCR process:
To improve OCR results, train the OCR process to recognize characters more consistently each time they are encountered. To train the OCR process, specify a training file that contains training information.
A training file matches patterns. The contents of a training file are special characters that are difficult to recognize during OCR. To improve conversion results, the reliable characters of a training file are compared with questionable characters in the input document.
The Standard training file that is included with DocuXplorer is initially empty. To place training information in it, you add to it interactively, during the OCR process. You can also create a new training file , and add to it during interactive training. In fact, it can be useful to build several training files, containing different information, for recognizing distinctive fonts or special characters in different categories of files.
Interactive training consists of specifying a training file, enabling interactive training , then monitoring the OCR process and seeing the words that are designated as questionable. When a highlighted questionable word appears, you can correct it and add it to the training file, and so continue to improve word recognition in subsequent OCR sessions.
After you have information in a training file, and you specify that file for training , training occurs each time you do OCR, whether or not you do interactive training. For example, when you OCR from the toolbar or from the Edit menu, with Copy As Text, training occurs automatically in the background, although data is not added to the training file. That is, when you specify a training file, the contents of the file are always used for comparison during OCR of the input document.