Copy Text from Image Context Menu

From Release Notes 1.49.120

  • Added “Copy Text From Image” entry to image context menu. (#27513)

This is not working for me. I read the issue report to see if there was something “special” I needed to do and didn’t find anything. Looks like when I right click an image, the context menu should have an option to copy text from image but it doesn’t.



Brave 1.49.120 Chromium: 111.0.5563.64 (Official Build) (64-bit)
Revision c710e93d5b63b7095afe8c2c17df34408078439d-refs/branch-heads/5563@{#995}
OS Windows 10 Version 22H2 (Build 19045.2604)

This feature would be really nice, especially when an image with text is posted in a language other than English. I’m wondering if it is broke or if I am doing something wrong.

That’s a MacOS only feature. That’s why.
I don’t know why they didn’t added it on the changelog to make it clear when it is obvious it would cause confusion.

if you see in the PR/Issue, you would have noticed it only has the label OS/macOS to specify it is not just for desktop but only for mac as well. But it also includes only the QA Pass-macOS label, which means it was only tested on it to make it clearer, but normal people wouldn’t research that so Brave should have added it even if the Issue name doesn’t include “MacOS only”.

And PR would tell you This PR extracts text from bitmap image came from renderer by using Vision framework from macOS SDK.

But for example, $match-case is only used in Regex rules but you don’t see ‘regex’ in the changelog, you only would see it in the issue/PR made for it, so people won’t know that either unless they research a little more, which makes it bad, since most users won’t research or now how to research.

It would be surely better if Brave team would specify things more in the changelog rather than just use the issues names and move on.

2 Likes

OK. I wish that was more apparent.

I saw the macOS tag but also saw the OS/desktop tag and figured it was for all. I did see the macOS only issue, but thought since that was from January they pulled in all desktop OS for release. The test cases only being on macOS didn’t register at all.

The first part I would do it’s the second part that gets me. :stuck_out_tongue:

Thanks Emi for the clarification. Take care. :slight_smile:

1 Like

Yeah it takes half a second to add in parenthesis with MacOS only on it to avoid confusion, since this is something so specific to a platform.

I rolled my eyes when I saw that changelog entry, because I know it sounds like a good feature to test, so it would be a disappointment for Linux and Windows, because Brave made it seem like it was for everyone, well even for Android users, although Android users still have google lens and yandex and other companies with features like that to extract text from images.

Exactly! I wanted to see this one in action! lol

hi @Chocoholic @anon57438784

do you know if it offline feature or it require internet?

and if it online is there privacy rules about it or how it work?

thanks and have a nice day both of you :slight_smile:

I haven’t a clue and it is not like I can test or anything. :joy:

I tried searching Brave Help Center and Brave GitHub Wiki for information but couldn’t find anything. Of course, as mentioned previously, I might not know how and where to search, especially on GitHub. Over a year of looking and that site still confuses me! Half the time I don’t know what I’m looking at. :rofl:

Related quote on finding things…

Hope Emi pops in with some input! Take care. :slightly_smiling_face:

1 Like

i do not have mac also and even did not update my brave yet :joy:

the example here mention image from search engine

but no more details if open image in brave will work or how it work and which language it support

github make it easy for search for commit message and words or code

this all the commit related to that feature

but it would be time consuming to read it :crazy_face:

hope that emi show up and maybe she know some doc resource

thanks for both of you and have a nice day :slight_smile:

1 Like

@justsomeone1 As mentioned, I did follow links. In any particular issue, the only tab I can grasp for the most part are the “conversations”. The “commits”, “checks”, and “file changes” are a nightmare.

I have actually used some information I found in the file changes tab and even posted that information on community topics. But as a general rule, I just don’t understand what I am looking at.

I do wish there was a step-by-step with pictures and links to relevant information “for dummies” on the various things you might see in an issue report. I have tried going through GitHub, and external documentation, but I get lost, overwhelmed, and confused (as usual). Although I think I have learned a little, just by desperately trying to help community members with issues, I know I haven’t even put a dent into understanding how GitHub works!

As an aside, have to mention one more thing that I found funny… looking at “conversations” on issue #27513, I see my profile picture in the last posted comment. I guess someone was testing or something… but still, was funny seeing it there. :slightly_smiling_face:

Sorry for the long-winded post. And I think this was probably off-topic for the most part, related but off-topic. I tend to do that! Ugh. lol

Take care. :slightly_smiling_face:

@justsomeone1 it is only available in Mac because This PR extracts text from bitmap image came from renderer by using Vision framework from macOS SDK.

That means it is using this SDK: https://developer.apple.com/documentation/vision/recognizing_text_in_images

Of course, I don’t care much about Apple or what they do, but must be easy to find more info or more user friendly information about it, than developer information.

But pretty much seems like is handled by MacOS and iOS, so I guess Apple for years has worked to make it decently enough, so it should be okay to use, they even say it can work offline but we have to guess ‘real-time’ might be better in many or most cases.

2 Likes

Will make sure we update the release notes to reflect this. Apologies for the confusion.

3 Likes

sorry for late replay @Chocoholic

then it’s your time to join the coding world :wink:

i notice it also and was wondring who that person who post your image :joy:

do not be sorry it was funny and it always nice to talk to you and share what we have

thanks a lot and you too

1 Like

sorry @anon57438784 for late replay and thanks a lot and i was lazy to read all those comment well done

sure online would be better from performance point of view but how privacy of data handled that should reveled if that how brave use it

and i notice that Mattches come to rescue us

thanks everyone and have a nice day :slight_smile:

@Chocoholic and @justsomeone1 just letting you know that today Brave Nightly 1.52.16 got this feature on Windows and seems to work okay. It has a language selector, or it seems people will be able to select it if needed, but even “en” copied this Spanish text okay, the only thing failing was the ä instead of á maybe for the font, so not a big deal to fix that little tiny mistake.

2 Likes

thanks a lot @anon57438784 hope it’s linux turn now :slight_smile:

and have a nice day all of you :slight_smile:

@anon57438784 Thanks Emi for the heads up. I’ll go check it out. Take care. :slight_smile:

Is it possible that developers add a checkbox or something which allows to remove line breaks from OCRd text?

image

Ideally it should be:

CRIME
Offences 2010 2015 …
Total 30,532 40,353 …