How to extract text from a screenshot on Windows 10
Optical Character Recognition has been around for years, and it’s improved so that photographs of printed text, e.g., from a book, can be read accurately. Most OCR apps target scanned documents or PDFs that do not allow users to copy text. That said, if you ever need to extract text from a screenshot, an OCR tool is what you need.
OneNote OCR tool
If you have OneNote 2016 on your Windows 10 PC, all you need to do to extract text from a screenshot is insert the screenshot into a note/page, right-click it, and select ‘Copy text from picture’ from the context menu. Newer versions of OneNote still have this feature, but it is cloud-based. What this means is that you will have to insert an image, wait a few minutes, and then the ‘Copy text from picture’ option will show up.
Once copied, you can paste it anywhere, including on the same OneNote page.
PhotoScan OCR
OneNote works great, but it makes you wait a while before you can extract text, and you have to insert a screenshot into a page first. The process is basically much longer than it needs to be. If you want a quicker solution, give PhotoScan a try. It’s a free app, but it has advanced features that you need to buy. For OCR, the free version is enough.
Install PhotoScan from the Microsoft Store.
Open the app and go through its brief tutorial. Click ‘Browse Photo’ and add the screenshot that you want to extract text from.
The app will automatically show you all the text it found in the image, and you can copy it to your clipboard and paste it anywhere you like.
The text will be read left to right, but the app won’t care for any ‘borders’ or ‘dividers’ that separate text. What this means is that text that is in an app window will not be differentiated from the text in the system tray. In the previous screenshot, the app has found and displayed text from the Command Prompt window and, it has also displayed text from the system tray, i.e., the language button, and the time and date. The time and date are not displayed together. Instead, the language button’s text is inserted between them.
If this is a problem for you, and you need the app to keep the text separate, you can crop an image into smaller parts so that each part contains only a specific text which the app will read. It’s a little time consuming but it’s easier than typing the text out yourself.
