Dear Hiring Manager,
I am very interested in your job dealing with photo OCR on vehicle license plates, but first let me introduce myself. I have a B.S. from UCSD in Cognitive Computation where I also minored in Computer Science. I have been almost exclusively programming in Python for the last 2 years using a plethora of machine learning tools, and that includes plenty of work with tensorflow and convolutional neural nets.
As I already hinted, I have a great amount of experience working on image detection, as I've taken a large interest in computer vision since my first deep learning course a couple of years back. I have worked on a numerous number of CNNs to perfect image detection, even obtaining high accuracy from relatively small networks.
In a most recent project I wrote a CNN to be fed 4 frames from a game I made in order to solve any level, you can view it in my portfolio. It should be noted that my work on freelancer.com itself is limited, but that's only because I just started here, I can assure you that I'm more than qualified for the job.
Anyway, your project should require these few steps (as similar OCR projects would):
1. Pre-processing: removing noise, converting to greyscale, etc.
2. Feature extraction: figuring out what parts of the image are distinguishing of a license plate
3. Text Classification: using the feature we extracted to train a neural network and tell us the text
I'd love to answer any other questions.
Thank you,
Adrian Herrmann