Have you ever needed to quickly copy text from an image into a document? Or tried transcribing presentation slides, only to find yourself manually typing each word? New to Windows 11, Snipping Tool‘s text action capabilities solve these headaches and more using artificial intelligence.
In this guide, I‘ll show you how to harness Snipping Tool‘s optical character recognition (OCR) to extract text from screenshots, images, pdfs, and other visual sources. We‘ll also dive into the machine learning driving this technology under the hood. Let‘s explore!
OCR Uses AI to Understand Text in Images
OCR leverages neural networks, a type of AI algorithm modeled after the human brain. These systems detect patterns from analyzing thousands of text image samples during training.
Over time, the algorithms have grown shockingly accurate:
- In 2020, Microsoft unveiled layoutLM – an OCR model achieving over 99% accuracy on scanned documents and complex images.
- Similar investments from Google, Amazon, and other tech giants have inflated OCR accuracy scores even higher since then.
This means you can point your smartphone camera at a restaurant menu, click scan, and receive the full typed out text within seconds!
Why Snipping Tool‘s Integration Matters
Previously, Windows users needed to install third-party apps to enjoy text scanning capabilities. But native integration into Snipping Tool brings this utility to billions of additional desktop users.
Specifically, Snipping Tool removes tedious barriers like:
- No more downloading standalone OCR software
- No app license fees or subscriptions required
- No emailing images to yourself to process text externally. Snipping Tool handles it locally in two clicks!
Plus you retain full control over your images rather than relying on cloud systems.
Step-by-Step: Extract Text from Images
Let‘s walk through utilizing text action with Snipping Tool:
- Open Snipping Tool via Start or Windows + Shift + S
- Take a Screenshot of your desired image/text area
- Select the OCR icon (🡐) in the toolbar
- Click and Drag to highlight the text section you need
- Copy the selection! The text is now on your clipboard.
And you‘re done! From here you can paste into documents, notes, emails and anything else.
Some additional tips:
- Prefer keyboard shortcuts? Use CNTL/CMD + C after highlighting text rather than the Copy menu option.
- For quick access, create a Snipping Tool shortcut on your task bar
- Screenshot an entire web page by scrolling first, then snipping
- Try snipping a email screenshot, and redacting the personal details with one click!
Let‘s contrast the experience of extracting a 3-sentence paragraph from an image:
| Method | Time | Steps |
|---|---|---|
| Manual Typing | ~60 secs | Highlight, switch windows, paste |
| Snipping Tool OCR | 5 secs | Clip image, highlight, copy text |
As you can see, Snipping Tool makes easy work of tedious transcription.
Limitations Persist Despite Accuracy Improvements
While text recognition models have achieved near human-level performance on clearly scanned documents, some limitations around image quality persist:
- Low resolution scans prevent distinguishing some characters
- Warped perspective from angled images causes confusion
- Small font sizes remain a challenge without zooming in
- Background textures/graphics increase likelihood of errors
Thankfully there are a few steps you can take to improve accuracy:
- Increase image size and clarity as much as possible
- Adjust skewed images to align text sections
- Running OCR analysis on the original file type instead of screenshots boosts performance as well
So for best results, point your camera directly at a document when capturing the image to process.
Microsoft Leads the AI Assistant Race
While competitors like Google pioneered early search and language understanding, Microsoft now dominates two key battlegrounds – productivity software and cloud infrastructure.
Controlling the Microsoft 365 suite owning Office and Windows OS provides tremendous advantage. Direct OS-level integration of text recognition and other AI functionality builds a moat against rivals.
And Microsoft Azure‘s enterprise cloud platform powers the machine learning behind the scenes. Major productivity gains attract business subscriptions revenues to Microsoft‘s cloud.
With advanced proprietary models like LayoutLM under the hood too, Microsoft‘s lead in unlocking AI to enhance knowledge work continues to widen. Competitors can only access external OCR APIs with limited customization options.
The Future Looks Bright!
Consumer excitement and business appetite for computer vision and text recognition applications continues to skyrocket. Beyond productivity software, surging industries like self-driving vehicles and IoT smart devices are poised to leverage OCR capabilities as input sensors.
Within Windows, Microsoft Chairman Brad Smith hints at plans to incorporate generative AI soon too. Imagine Snipping Tool intelligently filling in incomplete sentences from poor scans with reasonable guesses!
In the meantime, be sure to take your new text extraction superpowers in Snipping Tool for a spin. Let me know which creative use cases or efficiency boosts you uncover! Screenshot key information from slides, brochures, even product packaging and street signs using OCR built right into Windows.
