How to compare two PDF tables with different contents

Key points of this article

First determine whether the two PDFs are text versions or scanned copies. The scanned copies must be OCRed first, otherwise the text cannot be selected and compared.
It is easy to miss rows and columns when directly comparing PDF tables with the naked eye. It is more reliable to export the two tables to Excel and then compare them.
In Excel, use the formula to determine whether the corresponding cells are equal, and then use conditional formatting to highlight the difference to quickly locate the difference.
Unify table headers, row and column order, and number formats before comparison, otherwise layout differences will be misjudged as content differences.
After the results come out, you need to check the original PDF to make sure there are no lost lines, incorrectly merged cells, or misplaced numbers in the export.

If you have two PDFs in hand, each containing a table, such as two versions of the quotation, two months of financial statements, or the attachment table before and after the contract modification, the time-consuming thing is often not to find out "where has been changed", but to confirm whether each cell has changed. Scanning line by line with the naked eye, it is easy to miss a decimal point or a misplaced number.

This article is written for people who need to check PDF tables, especially readers who want to figure out how to compare different content between two PDF tables. After reading, you will know: first determine the file type, then turn the table into data that can be compared cell by cell, finally how to locate the differences, and how to check whether the results are missing.

The following method does not rely on a fixed tool. The core idea is: PDF itself is not suitable for frame-by-frame comparison. Converting it into tabular data for comparison is a safer way.

How to compare two PDF tables so as not to miss the differences in cells? One sentence conclusion

First convert both tables into editable tabular data, and then compare them cell by cell. This is a method that is not easy to miss.

your situation	Recommended practices	reason
Both are text version PDFs with neat tables.	Export to Excel and compare cell by cell	Data can be judged using formulas, not with the naked eye
One or both of them are scanned copies	Do OCR first, then export to Excel	The text in the scanned document cannot be directly selected and compared.
I just want to know what has been changed in the two documents as a whole.	First use the PDF comparison function to see the overall differences	It is suitable to locate the page and general scope of changes first.
The table is very large, and rows and columns are often added or removed.	Align rows by key column first, then compare	Misaligned row order will misjudge the same record as different
There are merged cells or multi-layer headers in the table	First manually check the table header, and then compare the data	The merge structure may not be completely consistent when exporting

Before you start: Confirm these points first

Determine whether the file is a scanned copy. Use the mouse to select a number in the table in the PDF. If it can be selected and highlighted, it means it is a text version; if it cannot be selected and it can only be framed into a picture, it means it is a scan or picture. OCR needs to be done first, otherwise the exported table may be empty or garbled.
Confirm whether the structures of the two tables are consistent. Compare the number of columns, header names and order of columns. If the structure has changed, you must first align the columns in the exported table, otherwise the cell-by-cell comparison is meaningless.
Find a key column that uniquely identifies the record. Such as order number, project number, employee number. Matching by key columns is much more reliable than comparing by row numbers. Even if rows are inserted or deleted in the middle, there will be no overall misalignment.
Unify the format and compare again. Whether numbers include thousandths, whether dates are written in the same way, and how many decimal places the amount retains, will all be considered differences in content. Unify first, then the results will have reference value.
Prepare the desktop client. The PDF comparison, OCR, export to Excel and other functions involved in this article are all in the OmniDoc desktop client, which can be downloaded from download.omnidoc.top, and then follow the steps below.

Steps to compare two PDF tables using OmniDoc

Open two PDFs separately. Load the two files to be compared in the desktop client, and first confirm that both files can be displayed normally and the table content is complete.
Use the comparison function to view overall changes. Find the option to compare two PDFs in the PDF editing function and select both files in turn. The function of this step is to first see which pages have changed and narrow the scope. It is not necessary to read the entire document frame by frame from the beginning.
Note down the discrepant page numbers and table positions. Write down the page numbers of the tables involved in the comparison results, and only process these pages later.
Do OCR on the scanned document first. If the pages where the table is located are scanned copies, perform OCR recognition on these pages so that the text can be selected and exported. After the recognition is completed, check a few numbers to see if they are recognized correctly.
Export the two PDF tables to Excel. Select Export to Excel format in the PDF editing function to export two files separately. It is recommended that the file name indicates the version, such as "Quotation_old version.xlsx" and "Quotation_new version.xlsx".
Place two tables side by side in Excel. You can place two worksheets adjacent to each other in the same workbook, or open two windows to view them side by side to facilitate column-by-column comparison.
Use formulas to judge each cell and then highlight the differences. Enter a formula similar to "=B2=Sheet2!B2" in the blank column, where Sheet2 is the name of the worksheet where the other sheet is located. If the result is FALSE, it is a different cell. Then use conditional formatting to mark these cells with eye-catching colors. If the order of rows in the two tables is different, sort by the key column first, or use a search function to match the same record, and then do this step.
Compile the list of differences. Record the position, old value and new value of each difference into a table, and mark the PDF page number. The list makes it easy to check back and communicate with others.

Check each item after completion

Check items	How to see	What to do if there is a problem
Number of rows and columns	Whether the number of rows and columns in the two exported tables is consistent with the original PDF	Check whether any cross-pages are truncated and return to the original PDF to fill in the missing parts.
Are the numbers misaligned?	Spot check amounts, decimal points, negative signs and total lines	Corrected against the original PDF; the numbers recognized by OCR need to be reviewed manually, and machine recognition may be wrong.
Merge cells	Whether the merged areas in the original table are split in Excel	Merge manually, or explain structural differences in diff list comments
Is the difference real?	See if the difference is just thousandths, date writing or spaces	Compare again after unifying the format to remove false differences caused by the format.
Is the list of differences complete?	Randomly select a few places to compare with the original PDF	If there are any omissions, narrow down the range and re-compare the pages.

📊 Core comparison

There are usually three ways to compare two PDF tables, each of which is suitable for different scenarios.

method	Suitable for the scene	limitations
Side-by-side comparison with the naked eye	The table is small and only a few changes are possible	When there is a lot of data, it is easy to miss rows and columns, and it is difficult to retain records.
PDF comparison function	First locate which pages of the two documents have changed	It cannot replace cell-by-cell verification and needs to be judged based on the exported table.
Export to Excel and compare using formulas	There are many rows and columns, requiring accuracy in each frame.	Need to do structural alignment and format unification first

In actual work, it is recommended to first use the comparison function to narrow the range, and then use the method of exporting Excel and adding formulas to confirm cell by cell.

❓ Frequently Asked Questions (FAQ)

1. Comparing two PDF tables, is it enough to see with the naked eye?

This is OK when the table has only a dozen rows and few changes. When there is a lot of data or when verification records need to be kept, it is easy to miss it with the naked eye. It is recommended to export it to Excel and then use formulas to judge.

2. Can scanned PDF tables be compared directly?

Usually OCR needs to be done first to turn the text in the image into content that can be selected and exported. The identified numbers must be checked against the original documents and cannot be used directly as accurate data.

3. The order of rows in the two tables is different. How to compare?

Do not compare directly by line number. First find a unique identifier in a column, such as a number, sort by it or use a search function to match the same record, and then compare column by column.

4. If the comparison result shows that there is no difference, can you rest assured?

uncertain. First make sure that both documents are complete versions, that the tables are not truncated, and that OCR does not misrecognize numbers before drawing any conclusions.

📝 Summary

How to compare two PDF tables with different content? The key is to first turn the table into data that can be judged cell by cell: use the comparison function to locate the changed page, export to Excel, use key columns to align the rows, and then use formulas to highlight the differences. Check the original PDF after each step, and the result will be more reliable.

Open OmniDoc and export the PDF content as an editable file