3 ms·

I have produced license plate recognition software, at least enough to extract the plate text for subsequent OCR'ing. Paul Graham/Ycombinator weren't intereste

by zandorg 19y ago

I have produced license plate recognition software, at least enough to extract the plate text for subsequent OCR'ing. Paul Graham/Ycombinator weren't interested, so I gave up on trying to sell it. In theory, you could setup a webcam in a back garden and scan all the traffic going past in realtime. I am now working on other software, which is a shame, as Pretext is pretty nifty. http://www.sanfransys.co.uk/ http://www.sanfransys.co.uk/ (not a US company)

rms 19y ago
Did you try talking to governmental and non-governmental agencies about your software?
- zandorg 19y ago
  No, I only talked to other OCR companies, who didn't reply. Maybe I should give it a try, thanks for the hint and the positive response.
abstractbill 19y ago
Surely this technology is fairly well-known and widely deployed these days (for example, to enforce the London congestion-charging zone). What did your software do over and above what's currently being used?
- zandorg 19y ago
  It runs easily on personal computers, and because the development costs were low, it's cheap. So I was trying to compete on price and accessibility.
gscott 19y ago
> Paul Graham/Ycombinator weren't interested, so I gave up on trying to sell it. They can't take every idea, even when they are good ones. I like your idea, if you could make it more of a turn key system. I believe the customer would want a photograph of the vehicle, the license plate (in plain text), and time/date, and statistics on how many autos an hour/day/month pass by to add extra value. It could be sold to gas stations or even city governments for deployment in high crime areas. It would compliment video tape / dvr systems. If you can get the price below $1,000 then it would be affordable to security conscious companies. It would have to be automated, and be a server so the company can login and view the data.

by zandorg 19y ago

rms 19y ago

Did you try talking to governmental and non-governmental agencies about your software?

zandorg 19y ago
No, I only talked to other OCR companies, who didn't reply. Maybe I should give it a try, thanks for the hint and the positive response.

abstractbill 19y ago

Surely this technology is fairly well-known and widely deployed these days (for example, to enforce the London congestion-charging zone). What did your software do over and above what's currently being used?

zandorg 19y ago
It runs easily on personal computers, and because the development costs were low, it's cheap. So I was trying to compete on price and accessibility.

gscott 19y ago

> Paul Graham/Ycombinator weren't interested, so I gave up on trying to sell it. They can't take every idea, even when they are good ones. I like your idea, if you could make it more of a turn key system. I believe the customer would want a photograph of the vehicle, the license plate (in plain text), and time/date, and statistics on how many autos an hour/day/month pass by to add extra value. It could be sold to gas stations or even city governments for deployment in high crime areas. It would compliment video tape / dvr systems. If you can get the price below $1,000 then it would be affordable to security conscious companies. It would have to be automated, and be a server so the company can login and view the data.