Any easy way to get a translation of Chinese from an online image that's not "text"?

Not sure if this belongs in Technology or Learning Chinese, but whatever: again and again I deal with Chinese websites that have Chinese in an “image”, just something that the web browser recognizes as a picture. There’s no text to read in this image, nothing I can select and copy/paste into Google Translate. (Right now, for example, my university has a big “image” like this telling us what’s going on with the Wuhan virus.)

So: is there a convenient way to somehow get this rendered as text? I use PDFPen, and that can give me OCR (Optical Character Recognition) for a whole bunch of languages, but not Chinese. Am I missing some other way to copy “text that’s actually part of a picture” into Google Translate or my iOS Pleco app?

FWIW, I’m in the MacOS / iOS ecosystem.

Thanks in advance!

Both google translate and pleco have OCR on your phone. I think you have to pay for pleco’s OCR though. If I remember correctly there’s also a way to upload a document (should include pictures) into the web version of google translate.
There are also other OCR packages that you can download and make your own python script if you’re coding-minded.

1 Like

Yea and sometimes it’s a poor scan/photo of the document where the text is barely readable or even worse, handwritings…

Maybe if you can like work it into a captcha so it can be typeset by a human…

1 Like

Scan it with your phone in Google Translate? Just as an odd example, worked fine on this

I noticed an Import feature on my phone as well. Maybe that’s what these guys are talking about :slight_smile:

4 Likes

Google Translate and Google Keep. There is no need to copy and paste. Just import the picture into translate and done. I import, read, translate Chinese scratch daily. Keep is kind of nice because it automatically creates a note to keep as long as needed.

2 Likes

你好! 我愛臺灣!

1 Like

Huh. Thanks all. On my phone it seems to work fairly easily - just hold up the phone to the image, and it gives me a translation. Pleco can give me a transcription (OK but not perfect accuracy, and apparently not that “smart”: for example, it scanned “2月10日” as “2月10臼”, which seems significantly less likely), and then I can import that onto the computer, if I so choose.

On my computer, dragging the image into Google Translate’s window doesn’t seem to do anything. I’d like to somehow see what I normally do with Google Translate - the Chinese on one side, the “English” on the other. Is there a way to do this that doesn’t involve 1) clipping the image on my computer, 2) magnifying the image, 3) scanning it with Pleco on my phone, 4) importing that scan back into the computer? That seems like a long way around!

You could try this on desktop Windows or Linux:

Based on Tesseract OCR. Should be able to handle Chinese.

1 Like

I’m in the Mac world, so unfortunately this one won’t work for me.

I said above I use PDFPen, which is software I quite like; PDFPenPro does additionally include Chinese (as well as other languages). Hm, base software is USD$125, but if I do an upgrade it’s USD$50 … doubt it’s worth it to me.

Maybe you have done this or maybe I’m not understanding.

I still haven’t seen where you said you tried the “import” function of Google Translate, where you’ll have Chinese, Pinyin, English after. Yes, have to grab a clip, screenshot, picture or something.

Google Keep will also do it and put it into a Google Keep Note.

Chrome automatically translates all Chinese web pages if desired but not “pictures”.

You mean on my computer? I think I’m trying that … I’ve got the “double pane” of Google Translate. I select “Documents”. I browse my computer and select a file. I import a PDF of Chinese text (that is NOT OCR’d: it’s just a picture - no recognized text to highlight). I select Translate. I see “Translating” for a few seconds, and then I see “Page 1” at the top of a totally blank page. That’s in Safari.

Quick hop into Chrome to try the same thing … nope, same result: “Page 1” at the top of a blank page.

Right now it seems like it’s relatively easy to handle these “images” on my phone or iPad, but not so easy on my computer.

Looks like it’s a phone app, not desktop.

iPad, but works the same.