Is it possible to reverse image search on my local computer?
I have a couple thousand photos in a folder, and I want to know if a specific photo is among these files. Of course, the file size and resolution and the checksum would differ, so it ain't easy without manually checking the photos 1 by 1.
>>54513521
upload them all to google drive and make them public.
then use image search
>>54513521
I have never done this before but this sounds like it can be done quite easily with machine learning. Since you are looking for a single image you can easily train any model with your single image. This will result in overfitting and thus make the model find this specific image. Some care still has to be taken that it doesn't overfit so it depends on dimensions and stuff but that shouldn't be too hard. Once you have your silly model, just classify all images and get the one with the high response. Obviously, this is not how google does it since you do a lot of computation per query but in your case something like this can work.
for(each(file in directory)):
If file ==== image {
Print(image) + return
}
end for!
Yes. Google something like 'duplicate image search' or something to that effect. Programs exist that use heuristics/needn't be an exact match. Use /sqt/ next time.
>>54513681
I actually found a promising software called VisiPics
http://www.visipics.info/index.php?title=Main_Page
DoubleKiller
>>54513521
Here is my idea.
Resize every photo to 128x128 or something.
Then compare them and order by difference. Probably YUV or something to compare.
Wouldn't work for cropped photos.
>>54513521
Awesome Duplicate Photo Finder on M$ Winblow
>>54513706
what language is this?
>>54513521updatedb
md5sum image-youre-looking-for.jpg
locate *.jpeg | xargs md5sum | grep MD5SUMOUTPUTGOESHERE
>>54515225
>Of course, the file size and resolution and the checksum would differ
>>54515279
my bad, too lazy to read
>>54513521
Why is /g/ so fucking stupid -_-
Google "Perceptual Hash"
>>54514562
You can't be serious... right?
