Automatic field completion

Simple fields defined as text boxes allow auto completion. This feature completes the user input for the current field by OCR results found on the current document page. For each inserted character, the input text is checked against the page OCR results. If less than 11 matches are found, a list is displayed. The matches are highlighted on the page in the Document Viewer, and the current list item is displayed in the image MiniViewer and marked with a red or green border in the Document Viewer. By default, the input text is completed with the first list item.

If the correct item is already in the list, you can use the Up and Down Arrow keys to select the item and press Enter to validate the field. The list may contain the same entry several times. This is because the same OCR results be may found several times on the document but at different locations. If you use specific knowledge base learning, make sure that you have selected the correct item from the list, as the geometrical information is needed to train the project correctly. If neither the mouse nor auto completion is used to fill in data in Validation, extraction online learning cannot reach the best level of performance.

If the selected entry in the list contains an OCR error you can press Esc to turn auto completion off. This enables you to edit the value without affecting the field coordinates that are displayed in the Document Viewer. If auto completion is NOT turned off and you edit the field, the field coordinates are not kept. As a result no highlighting is available in the Document Viewer for the field. If auto completion was turned off you can clear the field to enable it again. When you now type new characters into the field auto completion provides a new list of available OCR results.