this post was submitted on 06 Oct 2024
592 points (99.7% liked)
Technology
59287 readers
5759 users here now
This is a most excellent place for technology news and articles.
Our Rules
- Follow the lemmy.world rules.
- Only tech related content.
- Be excellent to each another!
- Mod approved content bots can post up to 10 articles per day.
- Threads asking for personal tech support may be deleted.
- Politics threads may be removed.
- No memes allowed as posts, OK to post as comments.
- Only approved bots from the list below, to ask if your bot can be added please contact us.
- Check for duplicates before posting, duplicates may be removed
Approved Bots
founded 1 year ago
MODERATORS
you are viewing a single comment's thread
view the rest of the comments
view the rest of the comments
I’m looking for some adversarial material - numbers and letters at various angles that I can stick to the left and right of my license plate. To a human it will be obvious which part is my license plate but it might be sufficient to confuse an ALPR algorithm.
The readers are smart enough to distinguish between them, so it won't actually do what you want. You could try to flood the plate with IR and cover the plate with clear-to-human IR reflecting cover. Might work. Might not.
Readers are not smart. They are trained on data with license plates, and I doubt their training had license plates with extra characters on both sides.
The reader to which I was referring, is the entire system, to include the server-side processing. If it's able to create a searchable db of political standings, we should assume it's able to trim excess characters.
However, I'm not telling you what to do at all. I don't know how they operate; I'm just making assumptions based on what they said and my knowledge of the processes used. Basically, just adding to the general knowledge pool, so someone smarter than I will have more data to make a more informed decision.
The article gave me the opposite impression. Basically their database contains lawn signs and bumper stickers on accident - they save all images where text is found but they keep it just in case it had a license plate (because they aren’t sure what is or isn’t a license plate). These kinds of databases are so massive there’s little to no human eyes on images. Anyway I don’t think it would be very hard to send garbage into their database.