How to adjust the format of the picture after it is converted to Word?

Key points of this article

If the format is messed up, it is not necessary to retransmit it. First determine whether it is a line break, font, table, or recognition error, and then decide how to correct it.
Each line is broken into short paragraphs. This is usually caused by the original image being shot in lines. This can be greatly improved by using search and replace to clear out redundant line breaks.
When the table is messed up, rebuild the table according to the number of rows and columns in the original image, which saves more time than hard-adjusting the cell lines.
Text recognition itself is not accurate, and it is useless no matter how good the format is. You should deal with the recognition quality first, and then adjust the format.
Numbers, dates, amounts and formulas must be checked item by item with the original image after conversion, as machine recognition may be wrong.

Many people convert a photo of courseware, a screenshot or scan of a paper contract into Word and want to modify the text directly. When I opened it, I found that the title ran into the middle of the text, the paragraphs were cut into line after line, the tables turned into scattered text, and the fonts and line spacing were also varied. How to adjust the format after the picture is converted to Word is messed up. This is the reason why many people get stuck at this step. This kind of problem often makes people mistakenly think that the conversion tool is not good. In fact, in most cases, the layout information of the original image has not been completely restored, and the converted text structure needs to be sorted manually.

This article is for people who have gotten pictures or scans and need to put the content into Word to continue editing. After reading, you can first determine what type of mess it is, and then troubleshoot and repair it in a fixed order. You will also know how to take photos and prepare materials in the future to reduce rework.

The format of the picture is messed up after being converted to Word. Do you want to convert it again? One sentence conclusion

Don't be in a hurry to redirect yet. In most cases, it is more time-consuming to repair existing documents one by one according to the type of problem than to re-convert them; only when a large area of text is recognized incorrectly, it is worthwhile to re-recognize it in another way.

your situation	Recommended practices	reason
Fonts, font sizes, and line spacing are not uniform	Set the format of a piece of text, and then use the style or format brush to apply it in batches	The text is basically correct, just needs to unify the appearance
Paragraph is cut off line by line	Use find and replace to remove excess line breaks, and then manually fill in the required segments	The end of each line of the original picture is often treated as the end of a paragraph.
The table may be misplaced or become ordinary text.	The table is rebuilt according to the number of rows and rows in the original drawing, and the complex form is sorted out by OCR first.	Table lines and cell information are easy to be lost in conversion.
Large area text recognition error	Re-identify it with a PDF editor with OCR function, and then export Word	The recognition quality is low, and the revision of the format is of little significance.
Long-term editing or multi-person collaboration is required	Import UniDoc web page editor (app.unidoc.top) to continue typesetting	It is compatible with Word editing for easy subsequent maintenance

First determine what kind of problem it is

See the symptoms clearly before making repairs to avoid making changes to all areas at the same time. The more you make changes, the more chaotic it becomes.

Symptom	Common reasons	How to confirm
Each line breaks and becomes a lot of short paragraphs.	The original picture is taken according to the line, and the line break is kept as a paragraph.	Open the paragraph mark to see if there are paragraph symbols at the end of each line
The font changes from large to small, with a mixture of Chinese and English fonts.	There is no uniform font information in recognition.	Select a few paragraphs to see if the font and size are the same
The title and page number are mixed into the middle of the text	The header, footer and body are identified on the same layer.	Check the header footer area and text reading order
Double columns are strung together.	The double-column layout has not been restored to separate columns.	Compare the original picture to see if the reading order jumps back and forth
The table becomes space-aligned text	Table lines are treated as ordinary graphics	Click on the table area to see if the cell can be selected
Formulas, special symbols become garbled or boxes	Missing font or unrecognizable symbol	Compare the original image to confirm whether the font supports the symbol.

# item-by-item troubleshooting and repair

Do it in the following order, easy first and difficult later. After each step, turn a few pages to confirm the effect, and then move on to the next step.

Overall browsing: use the page view to look through the pages and write down which pages and areas need to be repaired. Make a list first, and then change it, so as to avoid making changes at the same time.

two. Clean up excess line breaks: replace with search, remove the manual line breaks brought by the original image line by line, and then make up for the areas that really need to be segmented. This step usually solves more than half of the "line-by-line" problem.

Unify fonts and paragraphs: first select a paragraph, set the font, size, line spacing, and first line indentation, and then apply it to other paragraphs in batches with formatting brushes or styles.
Fixed title level: set the title style to the chapter title so that the catalog and navigation can be identified properly, and it is also easy to adjust later.
Handle headers, footers and page numbers: delete identified duplicate headers and reinsert page numbers as needed to prevent them from mixing into the body.
Restore columns and text boxes: double-column content is set with column-by-column function; scattered text boxes are checked one by one, and those that can be incorporated into the text are changed into ordinary paragraphs.
Rebuild the form: when the form is seriously confused, delete the wrong form, create a new form according to the number of rows in the original picture, and then fill in the contents one by one. Tables with a large number of rows and complex lines can be re-identified and then sorted out with OCR.
Check page by page: compare with the original picture, focus on numbers, units, formulas and proper nouns, confirm it and then save it.

what if the text recognition itself is not on time

If the original text is misidentified, it is useless to fix the format. The scanned version of PDF can use the PDF editor in the desktop client (download.omnidoc.top) to do OCR, identify it, export it to Word, and then enter the above troubleshooting process. Photos or screenshots can also be sorted into PDF and processed page by page, making it easier to control the results than direct conversion. When you need a translation, you can upload the file in omnidoc.top, choose bilingual control or translation only, and also check typesetting after download.

how to deal with the quality of the original image

Shooting skew, shadow, reflection or low resolution will make the layout clue lost, and the confusion after conversion often comes from this. In this case, a remake is more effective than a hard tune in Word.

# how to avoid the next occurrence

-when taking photos, the lens is facing the page, the light is uniform, shadows and reflections are avoided, and the text is clear. -the multi-page materials are sorted out according to the page number first, and then processed after they are uniformly saved as PDF. Do not convert them piece by piece. -think clearly about the purpose before switching: just the text, or keep the layout. As long as the text, the format can be simplified, saving a lot of adjustment. -complex tables and formulas are handled separately. Don't mix them with the text. -use a page or two to test first, confirm the effect, and then process the whole material in batches. -important content must be checked with the original picture after conversion. Machine recognition may be wrong and should not be used directly.

# frequently asked questions (FAQ)

1. The format of the picture is messed up after it is transferred to Word. Can it be restored to its original appearance?

It's hard to be exactly the same, but it's usually close to the original. The idea is to make sure the text is correct, and then revise the paragraphs, fonts and tables one by one.

2. Why is the format of the photos taken by mobile phone more messy after being transferred?

Tilting the shooting angle, shading or low resolution will make the layout clue lost. It is usually much better to retake a photo with uniform light facing the page.

3. What should I do after the table is converted?

Tables with few rows and rows can be rebuilt more quickly directly in Word; for those with many rows and complex lines, first use OCR to sort them out, and then check the contents one by one.

4. Do machine-recognized characters need to be checked?

Need. Figures, dates, amounts, proper nouns and formulas are prone to errors. It is recommended to check them with the original drawing item by item and confirm them before use.

# Summary

The format of the picture is messed up after being transferred to Word, which is mostly caused by the loss of line breaks, fonts, tables and layout clues. It is more efficient to judge the type of problem first, and then deal with it in the order of cleaning up line feeds, unifying format, repairing tables, and checking contents. Pay more attention to the next time you take photos and organize the materials. There will be much less rework after the pictures are transferred to Word.

Open OmniDoc to edit and export documents online