I frequently find images which I strongly suspect are copyrighted images stolen from the internet.
When this occurred, I used to reverse image search with Google to see if the image was used elsewhere on the internet. If it was, I would simply flag the image for copyright and move on.
But recently, Google removed the ability to search an image altogether, and now I can no longer do this. So how am I supposed to find an image’s source and tell if it is copyrighted? Is there no way to do this anymore?
[unrelated side note but why has Google been making all of their services so much worse recently? it’s awful]
My understanding is that the feature has been removed. It has been removed not only from the right click menu but also the bottom of the Google Lens sidebar. I was told on Reddit that the feature had recently been removed completely.
Personally I search with TinEye and use an extension for that with Firefox so I just double-click the image and search TinEye. You may want to consider switching to Firefox or another third party browser if you’re not into Google. I’ve been using Firefox as my daily browser for years now.
Perhaps I’m fortunate to have an outdated laptop with an old MacOS (10.11.6) and Chrome (103.0.5060.134). In Google Images, I can drag an image onto the search bar. It opens up Google Lens, and I can search for an Exact Match, giving me the source of those matches.
Yeah, google getting rid of their basic image search sucks, it makes things much more complicated. (Since you bring up the question of why google’s services are getting bad, you may be interested in this which explains a lot of it).
Google lens still picks up some of them, just not as many as the old search did. It’s still worth doing though. Also, it’ll often include the image somewhere in the results but not recognize it well enough to highlight it as an exact result.
I mostly use this browser extension: https://github.com/dessant/search-by-image
I have it set to search Google lens, Bing images, Yandex, Tineye, and Pinterest. Right-click an image and use the “all search engines” option, and it’ll open a new tab for each engine.
IDK, Google Image Search still seems to work for me (just tested it with Chrome and Firefox on Mac and with Chrome and MS Edge on Windows).
Copy an image URL to the clipboard.
Go to images.google.com (don’t use the first input box that appears)
Click the camera icon (“Search by Image”) and you’ll get another input box
Paste the URL into the box labeled “Paste image link”
Click Search
Google comes up with Google Lens results of variable utility (ignore these)
Click Find Image Source
Review the list of sites where the image was found
TinEye is also great. I find using both TinEye and Google Images is helpful if you can be bothered. Years ago Yandex had good reverse image search results, but I probably wouldn’t use that now.
For observations in South Korea from users who appear to be Korean, I’ve done a general image search on local search engine sites Daum and Naver using the Korean common name. There have been a couple of times when I’ve found evidence of copyright infringement that way, though I understand it’s a more complicated process than being able to drag-and-drop an image like with TinEye.
My impression is that the switch to Google Lens has made the results in step 7 less useful than in the past. The “find image source” now only seems to find exact matches. In the past, it would also often give some results for images that looked similar (instead of results for what it thinks it is). When checking whether a photo in an observation is copyrighted, I need it to be looking for fuzzy matches, too, because it is not unusual for users to take a cellphone photo of a computer screen, resulting in the image being cropped or tilted compared to the original.
Thanks for clarifying that. My previous experience was also that similar images used to be included in the Google Images results. Checking now, I do see that Google Images returns results with different sizes and formats, including slight color variations, but this may be a narrower set than before.
At a minimum, I’d like a reverse image search to return images with slight digital modifications (resizing, cropping, watermarks). TinEye is good at that, but my previous experience is that Google Images has a broader reach.
I’m a little less bothered about the more challenging scenario of a cellphone photo of a computer screen. In my experience, the source material for a fair proportion of those photos-of-photos was created by the same user (person takes a bunch of photos with camera, uploads to computer, takes photos-of-photos to get iNat IDs.)
