What to set up before searching
ImageTextFinder does not read every website immediately after installation. In onboarding or the extension status screen, you choose whether to allow all regular websites or only the current site.
Without access—or when Ctrl+F replacement is off—Chrome’s normal Find opens. Chrome Find also remains available on restricted pages such as Chrome internal pages, the Chrome Web Store, and the built-in PDF Viewer.
Search page text and image text together
- Open a regular HTTP or HTTPS webpage and wait for the images you need to load.
- If this is a new site, open the toolbar status view and enable access for the current site. Reload a page that was already open when permission changed.
- Press Ctrl+F on Windows or Linux, or Command+F on macOS. You can also use the configured fallback shortcut.
- Enter a word or phrase in the combined finder. Visible HTML matches appear first, and image matches join the list as local OCR completes.
- Use Enter or the next control to move forward. Use Shift+Enter or the previous control to move back.
- Choose an image result to scroll to the source. When coordinates are reliable, the matching area is highlighted inside the image.
Search while OCR is still running
You do not need to wait for every image. Ordinary page matches and completed image matches remain usable while the extension progressively processes images near the viewport and then continues through the page.
Stop pending image work when you no longer need more results. Closing the finder cancels remaining work for that session, and retry starts only failed images instead of repeating every completed item.
When expected image text is missing
Check the image and language settings
Confirm the image has loaded and choose the OCR group that includes its printed language. English and Korean are enabled by default; other languages may require an additional packaged group.
Image source-host access
If Chrome cannot expose pixels from an image hosted elsewhere, access to that exact source host may be required. ImageTextFinder does not follow redirects or bypass access controls.
Unsupported content
Canvas, video, WebGL, DRM content, closed shadow roots, inaccessible cross-origin frames, and PDF Viewer are unsupported or not guaranteed. Recognition also varies with resolution, contrast, font, rotation, and layout.