The task of manual editing of scanned and recognized text of a book is required by clients who want to digitize any scientific work or work of art that exist only in printed form. As a rule, these are books, magazines, brochures of a very respectable age with damage and unclear printing. Automatic text recognition copes with such work, but the resulting text contains a very high percentage of errors, typos, and sometimes distorted information. For scientific works and books, this can have very unpleasant consequences. To get the best result, instead of automatic text recognition, you need to involve an editor who will proofread and correct the text in accordance with the original document.
After scanning or photographing a document, an image of its pages is obtained in raster format. This means that visually the scan is similar to a text page, but you cannot select and copy the text, edit it, or use the document search. To solve these problems, you need to perform text recognition. The result of the work, as a rule, is a file in the DOC, XLS or text PDF format not converted to curves.
Recognition of scanned text and editing is performed with the participation of the editor. He configures the program, makes adjustments during the recognition process, checks recognition errors and corrects them. The result is saved in the text editor. The output is a structured, editable, high-quality document in the required format (Word, Excel).
What factors can affect the cost of editing recognized text?
Poor quality page scans - increase the time of work and the cost of editing.
Editing recognized text with pictures takes longer, but does not always require additional payment.
Editing text with the recreation of the table structure depends on how the automatic recognition program builds tables.
Editing scanned handwritten text is a separate service, all signatures, stamps, notes made by hand must be re-typed by the editor, automatic recognition is useless in this case.
