# Table image to CSV

> Recognize table text and infer CSV cells from spacing.

[Open tool](https://www.toolcabana.com/su/table-image-to-csv) · [PDF and documents](https://www.toolcabana.com/su/category/pdf-and-documents)

Tool ID: table-image-to-csv. Requested language: su. Description language: en. Complete guide translation: no; untranslated sections use English.

## Overview (en)

Table image to CSV recognizes the text in a photo or screenshot of a table with Tesseract OCR and arranges it into CSV rows and columns. A new cell starts wherever the horizontal gap between words is wide, and words below a minimum confidence, 30 by default, are dropped. The result is shown as a table and downloads as table.csv.

## Supported tasks (en)

Browser Tesseract OCR downloads language data on first run. Table columns are inferred from word spacing; invoice/receipt fields use patterns. Review every extracted value before use.

## Steps (en)

1. Choose a source file. If the tool accepts multiple files, arrange them in the order you want.
2. Set ocr language, minimum ocr confidence.
3. Run table image to csv, review the output, then use the available copy or download controls.

## Settings

- OCR language (en; key: language; type: text)
- Minimum OCR confidence (en; key: confidence; type: number); minimum: 0; maximum: 100

## Limitations (en)

Browser file and text limits apply, along with the bounds shown in Settings. Image processing is capped at 32 megapixels. OCR quality depends on source resolution and layout.

## Example input

```text
A useful tool makes everyday work easier. Review the results before sharing.
```

## Privacy and connections (en)

OCR runs in this browser. The initial language-model download contacts a public CDN; the source image is not uploaded there.

## Questions (en)

### How does the tool decide where table columns are?

Columns are inferred from spacing. Words on the same OCR line are sorted left to right, and a gap of more than 25 pixels starts a new cell. Merged cells and complex grids may need manual correction.

### What should I do if the table image gives no confident cells?

Lower the minimum OCR confidence setting, which ranges from 0 to 100, or use a sharper, higher-resolution image. Words recognized below the threshold are left out of the CSV.

### Can I OCR a table that mixes two languages?

Yes. Combine up to four three-letter language codes with plus signs, such as eng+spa. The OCR language data is downloaded when you run the tool, while the image stays in your browser.

### Where is my input processed?

OCR runs in this browser. The initial language-model download contacts a public CDN; the source image is not uploaded there.

### What are the input limits?

Browser file and text limits apply, along with the bounds shown in Settings. Image processing is capped at 32 megapixels. OCR quality depends on source resolution and layout.

## Related tools

- [PDF to Markdown](https://www.toolcabana.com/su/pdf-to-markdown)
- [Document to structured JSON](https://www.toolcabana.com/su/document-to-json)
- [Invoice field extractor](https://www.toolcabana.com/su/invoice-extractor)
- [Receipt to spreadsheet](https://www.toolcabana.com/su/receipt-extractor)