Prepare to be underwhelmed.

Working With OCR

In which I try to reason with my phone.

This week, we were supposed to add an OCR app to our computer or phone — easy, I thought. I can add apps. I add apps all the time. Mostly games or chore reminders, sure. Because I am a fun person. But I know how apps work.

I was wrong.

Trying to use Google Keep.

Ok, maybe not completely wrong, but my first try was unsuccessful. My first try with an OCR app did not work, therefore it was a disaster.

I tried Google Keep first, because I use other Google applications a lot (mostly for convenience; I am not sure if I genuinely like it or just haven’t tried anything else better yet). After I installed the Google Keep app, I tried taking a photo of a page from a gardening magazine I subscribe to (further proof that I am really fun) and uploading it to my Google Keep app. I found the “Grab Image Text” option, clicked it, but got a little pop up saying “Can’t recognize text.” I tried again, this time using a screenshot of a page about flower gardening I found online — still “Can’t recognize text.” So I had to find a different app. Oh no.

My second choice was Evernote Scannable, because I have heard good things about Evernote and have been considering the digital planner offered by the company (also the logo is cute), but it is apparently lacking via Android, according to the article linked for the class, so I chose Text Fairy. Honestly, most of my decision came down to the fact I really liked the app name, and the lil’ dude who is the mascot. (The internet, with all of its shiny things is probably a dangerous place for me.)

I found Text Fairy immediately less frustrating, in that it actually seemed to work. I tried it with both a photo I took and a screenshot. Understandably, the app had an easier time reading the screenshot’s words versus my those from my best attempt at a still image of a page. A notable downside of taking a photo of a page of text, rather than being able to screenshot it, is the fact one has to contend with shaky hands or light reflections that make text harder to read. I honestly didn’t try very hard to get the photo perfect, because I was curious to see how well the app did at picking up text.

From the Text Fairy App.

The app did pretty ok. It did not produce a very neat selection of the text from the photo, which was disappointing but not surprising. The spacing was all over the place. However, on the app itself, I could copy the text and paste it elsewhere, which for research and essay-writing purposes is great.

The only real issue I had was at the PDF stage — it seems that, instead of just saving the text as a PDF, the app instead converted my entire photo or screenshot into a PDF document. I am not sure if this was supposed to happen or not (was it supposed to make a PDF of text only?), but I uploaded it to my Omeka page anyway, just in case! When I tried to “Send Text,” thinking it might be the solution to saving only the text, I got a “Document contains no text” popup.

One solution could be to copy-paste the text to a Microsoft Word document, and then save as a PDF from there, but it does add that extra step. When I write my essays, however, the copyable text is the most useful aspect; even if the PDF exporting does not work properly, being able to take a photo of a book page and digitize it so I can pull quotes without having to type the page out myself is a huge asset. However, I would still doublecheck before finalizing my essay, because illegible words get carried over from print to text incorrectly.

My image, and subsequently my PDF.

My PDF of the gardening magazine page, the thrilling conclusion to this nail-biting saga of my ineptitude with technology, can be found here: https://dragons.tierneytrieshumaningdigitally.com/cms/items/show/9.


Comments

Leave a Reply

Your email address will not be published. Required fields are marked *