Created by on May 06, 2016. Introduction Over the past few weeks I’ve been dabbling with deep learning, in particular. One standout paper from recent times is Google’s. This paper describes a system for extracting house numbers from street view imagery using a single end-to-end neural network. Sonic Youth Rather Ripped Raritan there. The authors then go on to explain how the same network can be applied to breaking Google’s own CAPTCHA system with human-level accuracy.

In order to get some hands-on experience with implementing neural networks I decided I’d design a system to solve a similar problem: Automated number plate recognition (automated license plate recognition if you’re in the US). My reasons for doing this are three-fold: • I should be able to use the same (or a similar) network architecture as the Google paper: The Google architecture was shown to work equally well at solving CAPTCHAs, as such it’s reasonable to assume that it’d perform well on reading number plates too. Having a known good network architecture will greatly simplify things as I learn the ropes of CNNs.

• I can easily generate training data. One of the major issues with training neural networks is the requirement for lots of labelled training data.

1>C: Program Files (x86) Microsoft Visual Studio 2017 Community Common7 IDE VC VCTargets Microsoft. This site uses cookies for analytics, personalized content and ads. By continuing to browse this site, you agree to this use. CodePlex is going read-only starting November 27th. Learn more and find next steps. CodePlexProject Hosting for Open Source Software. Register Sign In.

Hundreds of thousands of labelled training images are often required to properly train a network. Fortunately, the relevant uniformity of UK number plates means I can synthesize training data. Traditional ANPR systems hand-written algorithms for plate localization, normalization, segmentation, character recognition etc. As such these systems tend to be many thousands of lines long. It’d be interesting to see how good a system I can develop with minimal domain-specific knowledge with a relatively small amount of code. For this project I’ve used Python,, and.