OmniDocHow to convert scanned documents into Word and retain the original format
The scanned copy is a picture and cannot be edited directly as text. This article explains how to first use OCR to recognize text, then export it as an editable Word document, and check paragraphs, tables, and formulas according to the checklist to retain the original format as much as possible.
Key points of this article
- The scanned document is essentially a picture. You must first recognize the text through OCR and then export it to Word. These two steps cannot be omitted.
- Exporting is a step after recognition, and translation is two different things: translation to change language, and exporting to editable format.
- The more complex the layout, the more it needs to be checked after exporting. Make sure the text is accurate before adjusting the layout.
- Machine recognition will inevitably make mistakes. Numbers, names and proper nouns must be checked one by one against the original page.
- Tables, formulas and pictures are easily deformed, so they must be checked before and after exporting and processed separately if necessary.
You have a scanned PDF in your hand: it may be a document taken from a paper contract, it may be an old report, or it may be a page from a paper taken with a mobile phone. You want to change a few paragraphs of text, but find that the mouse cannot select it, the copied text is garbled, or the paragraphs are crowded together after being directly converted to Word, and the table is misaligned. This situation is very common because the scan saves a full-page image and what the software sees is not text.
This article is for those who need to turn scanned documents into editable Word documents. After reading, you will know what to judge first, how to complete the identification and export using the OmniDoc desktop client, and what areas must be checked after exporting.
One thing needs to be explained first: converting scanned documents into Word and translating documents into another language are two different needs. The purpose of conversion is to obtain a document that can continue to be edited, and the purpose of translation is to read it in another language. The two can be done one after the other, but there is no need to translate first in order to change the words.
How to convert scanned documents into Word so that the original format can be retained? One sentence conclusion
Direct answer: The scanned document must first recognize the text through OCR, then export it as a Word document, and finally check it on a page-by-page basis. The simpler the layout, the better the retention effect; the more complex the layout, the more time needs to be reserved for manual adjustments.
| your situation | Recommended practices | reason |
|---|---|---|
| Pure text contracts and letters, you just want to modify the text | First OCR recognition, then export to Word, and finally change the words | Mainly text, simple layout, few changes after exporting |
| Scan report with table | After OCR, export to Word and check the tables one by one; when there are many tables, you can export them to Excel at the same time for comparison. | Table lines and column widths are easily misaligned and require manual correction. |
| Scanned copy of the paper containing formulas and charts | Check formulas and picture positions before and after exporting. Important formulas can be inserted after taking screenshots. | Formula recognition is unstable and manual verification is more reliable |
| Magazines or papers with multi-column formatting | Check reading order after export and re-paragraph in Word if necessary | Recognition is carried out in visual order, and multi-column contents are easily serialized |
| I just want to understand the content without changing the original text. | Use the translation function and choose bilingual comparison or translation only | The purpose of translation and export are different. There is no need to convert to Word for reading. |
Before you start: Confirm these points first
1. Confirm that the document is indeed a scanned copy. Open the PDF and try to select a piece of text with the mouse. If it cannot be selected, or if only an image is selected, it means it is a scan or image PDF. If it can be selected normally, it means that it has a text layer and can be exported directly without doing OCR.
2. Check the image clarity. Recognition errors will increase significantly for pages that are skewed, blurred, shadowed, or wrinkled. When taking photos, try to keep them flat, facing each other, and with even lighting. For pages with poor shooting quality, it is recommended to rescan them first and then process them.
3. Figure out what to keep. If you only need text content, there is no need to worry about fonts and spacing; if you need a layout close to the original, you must prepare a budget in advance to spend more time checking. The goals are different and the focus of the inspection is also different.
4. Evaluate page count and file size. Documents of several pages can be processed at one time; documents of dozens or even hundreds of pages are recommended to be broken into several paragraphs and identified and exported piece by piece. This way you only have to rework a small section if something goes wrong.
- 5. Keep the original. * the scanned copy is an important original voucher. Save another copy before conversion. The exported Word is only an edited version, and the modified content should not overwrite the original.
# steps to convert scanned parts to Word in OmniDoc desktop client
- Open the download.omnidoc.top download OmniDoc desktop client, which supports Windows, macOS and Linux, and follow the prompts to install.
two. Start the client and open the scanner PDF you want to convert.
- Perform OCR recognition on the scanned piece to make the text on the page selectable and searchable. Recognition takes some time, and the more pages you have, the longer you wait.
- Spot check the recognition results: randomly select several paragraphs of text and compare them side by side with the original page to see if there are any wrong words, missing words or numerical errors.
- If necessary, modify the paragraph text directly in the PDF editor to correct obvious recognition errors before moving on to the next step.
- Choose to export to Word format and save as a new .docx file with a file name separate from the original.
- Open the exported file with Word, browse from the first page, and check the title level, paragraphs, tables, and headers and footers.
- Save the modified Word file with the original scanned document without any changes.
# check item by item after completion
Export is not equal to completion. The following table lists several areas that need to be checked, and it is recommended to go through them one by one.
| Check item | What do you think? | What if there's a problem? |
|---|---|---|
| Text accuracy | Spot-check numbers, names, place names and proper nouns against the original page | Modify it directly in Word; check important contents word by word, and machine recognition will inevitably make mistakes. |
| Paragraph order | Read through from the beginning and pay attention to whether the multi-column content is serial. | Cut, paste, and reposition paragraphs in Word |
| Title and hierarchy | Whether the title of the chapter is consistent with the original, and whether the font size is uniform. | Set it uniformly with the style of Word, instead of manually adjusting it segment by segment. |
| Form | Whether the number of rows, columns, and merged cells are correct | If the form is seriously misplaced, remake it in Word, or check it after exporting Excel. |
| Formulas and pictures | Is it missing, deformed, or misplaced? | Insert important formulas and pictures after screenshots, and mark the original page number |
| Number of pages and header and footer | Whether the number of pages is the same as the original, and whether there are any extra blank pages | Delete the blank page and fill in the header and footer according to the original |
When checking, give priority to numbers, amounts, dates, proper nouns and forms. The cost of making mistakes in these places is high and it is difficult to detect them at a glance.
# frequently asked questions (FAQ)
1. After the scanned piece is converted to Word, the format is all messed up, what should I do?
Don't be in a hurry to adjust the layout. Scanned pieces with complex layout are difficult to restore at one time. It is recommended to ensure that the text content is correct first, and then use the Word style to unify the title and text font size. Where you need to check, keep the original PDF open and watch.
2. Will the converted words be the same as the original?
Not necessarily. Fonts may be replaced with similar fonts in the system, and there may be differences in line spacing and spacing. If you have rigid requirements for fonts, you can uniformly format fonts and paragraphs in Word.
3. If you just want to translate the scanned piece into another language, do you want to convert it to Word first?
I don't need it. You can upload the scanned PDF directly in omnidoc.top, choose bilingual control or translation only, and download the results. Translation results also need to be checked, especially technical terms and figures.
4. Can the photos taken by the mobile phone be directly converted to Word?
It is recommended that you first organize the photos into PDF, and then follow the steps in this article. When shooting, keep the page flat, well-lit, do not tilt the picture, the recognition effect will be significantly more stable.
5. If there are a lot of scanned pieces of forms, why not export Excel directly?
If the core of the file is data, exporting the Excel makes it easier to calculate and filter. If the form is only part of the document, it is still recommended to export Word, and then check the contents of the table in the article. The two formats can be used in comparison.
# Summary
The key to convert the scanned piece into Word document and retain the original format is to identify, export and check item by item. First determine whether the file is scanned or not, and then check the text, tables and formulas according to the list to control the rework within an acceptable range as far as possible.
