LabVIEW
license plate recognition
thanks for your reply. I found one and edited it but it's still not perfectly readable. Can you give me some suggestions and comments?
I can give you a recommendation:
Clean up that block diagram! This is more than just cosmetic. A sloppy block diagram hides a lot of bugs that would otherwise be painfully obvious if better organized. This is the same reason why you have a structure in a text-based language. If you made random indents in your text-based code, the compiler wouldn't care, but it would be next to impossible to read and debug. The other thing I noticed is "sloppy code equals undisciplined thought process." And there is strong evidence that either one can affect the other. That is, creating clean code can lead to a more disciplined thought process, and the other way 'round. If your block diagram looks better after using the cleanup tool, it's probably time to review your coding practices.
Time constraints are a huge enemy of good code. But to let it drive your coding style is always counterproductive. Bad code performs badly, is difficult to impossible to debug and ultimately will always lead to longer development time.
So if your time constraints look to tight to do a good job, they are to tight to do the job at all!
Another thing about OCR (and any sort of image feature recognition) is the image quality. Lighting and good camera settings are the single most important aspects of image acquisition for such tasks. If the image is blurry, has bad contrast or is unequally lighted, you can throw weeks and weeks of development time at the project to make up for that in software, but ultimately you will always have samples that will elude your algorithme one way or the other and still require your computer to do huge analysis work for each and every image.
