These days, with the increasingly more powerful and capable cameras and sensors on our phones, mobile photography has virtually replaced a dedicated camera for most of us. Yes, those who are serious about photography still carry their DSLRs around, but for a quick point and shoot experience, or just to capture a single moment, a phone’s camera has become the go-to solution. However, have you ever considered that, more than just taking pretty pictures, your camera is now detailed enough to recognize individual objects, or even text? That’s what we’ll be taking a look at today, by teaching you how to use your camera to take a picture of a document and get the text recognized, sent back to you in editable format.
The technology that allows this is called OCR, Optical Character Recognition, and while it isn’t exactly new, it has certainly gotten better with the advent of better camera sensors and better recognition algorithms. There are also a number of apps that implement this sort of technology on Android – for example, Google Translate has an instant translation feature that allows you to see text translated in real time, using your camera, and it uses the same principle. However, today we’ll be taking about getting text from documents and other sources, and the app that we’ll be using for this purpose is called Text Fairy, which you can get directly from the Play Store, by using this link.
When you open the app for the first time, depending on your version of Android, you may also have to allow certain permissions, such as accessing storage. The app gives you a short intro about what it is able to scan, but you can proceed to get to the app’s main screen:
As you can see, the app has two main modes – you can either use it to take a picture and then analyze it, or give it a picture that you have taken beforehand to extract text. The app also lets you know which languages it supports, and if you tap the “Download your language” button, you can download additional idiom packs:
Tapping on the icon gives you a little rundown of what you can do in the interface:
And pressing the menu icon or swiping to the right gives you a few extra options:
For example, the tips section lets you know how to take a good picture of what you want to scan:
Anyway, enough preamble. Let’s try to capture a picture from the app by pressing the camera icon in the main interface, which will take you to the regular camera view:
Then, just find the document or graphic that you would like to translate. For example, here, I’m using the back of a book. Make sure the text is in focus and take your picture. If the picture you took isn’t good quality, the app will detect that and ask you to try again:
However, for a good picture, it will take you to this screen, where you can crop the selection and isolate the text portion by moving the blue circles around. When you’re done, press the arrow icon at the bottom:
Then, the app will ask you how many columns of text there are. For our purposes, there is only one, but it can also detect multiple columns. You can also pick the language here, using the dropdown menu:
You will see the app then process the image:
And then analyzes it in search of text:
Once that’s finished, the app will report the result, and you can copy it to your clipboard, share it, or save it as a PDF so you can use it on your computer, for example:
At the final step, you will see the text that the app extracted, and you can edit it, have it read back to you using Android’s Text-to-Speech capabilities, or simply paste it somewhere else:
The text will then be saved on the app, allowing you to come back later and use it if necessary:
For example, here’s a PDF of the text on my desktop computer:
That’s about it for today’s lesson. As you can see, this app is particularly good at recognizing text, and from my experience, it works much better than the alternatives. While it works best with printed text, like books, documents and magazines, even if it fails at recognizing your handwritten content, it can always serve as a way to store your notes in one place and make them more legible by processing them after the fact and allowing you to print them later. Any questions or doubts? Feel free to let us know using the comment section below!