There are times when you might see some on-screen text and want to grab it or use it in a document, only to find that for some reason it refuses to be clipped (e.g. it might be graphics-based or the document might be protected or whatever). A quick and easy solution would be a program that can clip the screen as an image and perform quick OCR (optical character recognition), and that, precisely, is what free program Capture2Text does, for more than 30+ possible languages.

Capture2Text has another component that can perform speech-to-text on short, ad-hoc voice recordings and convert them to useable text.

OCR languages supported: Capture2Text uses the open source Tesseract OCR engine. It supports the following languages by default: English, French, Spanish, German, Chinese, and Japanese, but a total of 33 languages can be added (including Bulgarian, Hungarian, Russian, Catalan, Indonesian, Serbian, Italian, Slovakian, Czech, Slovenian, Danish, Korean, Dutch, Latvian, Swedish, Lithuanian, Tagalog, Finnish, Norwegian, Turkish, Polish, Ukrainian, Portuguese, Vietnamese, Greek, and Romanian).

How to use: OCR

1. Download and unzip. The program does not require install. Run Capture2Text.exe
2. Right click on the Capture2Text system tray icon, go to settings then OCR language and choose the language you want.
3. Next, make sure the text you want is visible on screen. Place the mouse cursor such that the bottom of the mouse pointer just above the edge of the block of text you want to clip. Do this before you activate Capture2Text.
4. Use Win+Q on your keyboard, which is the default hotkey, to start the clipping process. Stretch the colored rectangular area that appears over the desired text, then click the mouse button to clip. You will see the OCR result displayed on the top edge of the screen (see screenshot above).
5. The OCR process will be performed and the results dumped into the clipboard by default. Go to your text editor, word processor, or wherever you want to use the text and paste it there.
6. Check to see if the text is sufficiently accurate. Some words may not be OCR’d correctly and might require manual intervention.

Note that a number of things are customizable, such as where to receive the resulting output, the color and opacity of the OCR clipping box, etc. To edit the default hotkey itself (WIn+Q) and other settings you need to edit the .INI file in the Capture2Text folder though.

How to use: speech to text

1. Press Win+4 to scroll through the supported languages (English and Japanese at the time of this writing)
2. Press Win+A to start recording. A message will flash on screen that reads “recording”.
3. Speak into the microphone. Once you are done (or a period of silence has been detected) the on-screen message will read “Analyzing”. If you would like to change the length of said period of silence, you need to edit the .INI file in the Capture2Text folder.
4. A dialog should appear with the various interpretations of what you just said. In the screenshot to the right, you can see the results I got when speaking “The rain in Spain falls mainly on the plain”. Select the most appropriate result.
5. That’s it, the text should now be in your clipboard (or whatever location you chose in the settings), ready to be pasted.

The verdict: this is the third program of it’s kind that we’ve written about at Freewaregenius, but is the first that is 100% free without caveats (ABBYY screenshot Reader was featured as a free giveaway which actually may still be available as you read this, and JOCR, another similar program, requires that MS Office be installed on your machine in order to run).

In terms of OCR performance Capture2Text is impressive (it OCR’d the images/text I threw at with a high degree of accuracy), and very well suited for your occasional, ad-hoc screen OCR needs (and doesn’t even require installation). The interface could have been more user friendly, etc, but c’est la vie.

I am less crazy about the text-to-speech component. I may be wrong about this, but it seems to be inferior to the Windows’ built in speech-to-text. If speech-to-text is what you seek, check out previously reviewed DicationPro speech-to-text.

If your OCR needs are more extensive than the occasional screenshot, check out our article titled “How to extract text from images: a comparison of 10 free OCR tools”.

[Thanks to reader Panzer for the tip about this software].

Version tested: 1.10

Compatibility: WinAll.

Go to the program page to download the latest version (~35.9 megs).

from freewaregenius.com March 06, 2012 at 06:23AM