Yeah, if I wanted to do a license-plate scanning startup -- I would just use an off-the-shelf dash cam and a laptop. No need at all for special optical equipment.
Under ideal conditions you'll be able to read a limited amount right in front of the camera. Want to read plates at/after dusk or to the sides (e.g. parked cars) and you'll need special hardware.
I have a fixed APNR camera on my property and even with good positioning it can't always read the plates (nor can I manually) because sometimes sunlight glare blows out enough of the image to hide a digit or two.
I promise you that the people building these systems didn't just fail to think of using an off the shelf camera before they went out and deployed the fancy stuff.
And you'd miss a lot of the cars. By hanging them off the sides of the car, they can scan both sides of the aisle in one pass. They also need to be weather proofed, hence the housings.
Depends what the company’s focus is, right?
If you want to do something with the data, why not use the already made, high quality devices that are relatively cheap. Seems like the important piece is likely using the data.
(Unless you are trying to create the devices themselves, of course)
I have a fixed APNR camera on my property and even with good positioning it can't always read the plates (nor can I manually) because sometimes sunlight glare blows out enough of the image to hide a digit or two.
I promise you that the people building these systems didn't just fail to think of using an off the shelf camera before they went out and deployed the fancy stuff.