1
$\begingroup$

I just started learning about machine learning recently and have a project where I have to develop a program for QR code localization so that a QR code can be detected and read at any angle of rotation. Development will be done in Python.

The plan is to gather various images of the QR codes at different angles with different backgrounds. From this I would like to create a dataset for training with neural networks and then testing.

The issue that I'm having is that I can't seem to figure out a correct feature design for the dataset and how to identify the QR code from the images for feature processing. Would I use ground-truth images or edge magnitude maps?

Any help with this would be amazing? Thanks for your time.

$\endgroup$
1
  • $\begingroup$ Just to understand the use case, given the image of a QR code, taken at an angle, you want to map it to some other image of the same QR code, taken at a better angle right? I mean the focus is on localization or QR codes? If localization, then you can use any existing datasets. For example, ImageNet challenge also has a localization challenge from where you can get your dataset and benchmarks as well. $\endgroup$ – Shagun Sodhani Jun 11 '16 at 14:13
1
$\begingroup$

The idea would be to find

1) Rotational independent and 2) Shift invariant features.

Simply speaking the Fourier transform is a rotational and translational (shift) independent feature. There are SIFT features used in computer vision that can extract such features. This is something you can try.

Since you are using QR codes,I think edge magnitude maps would be better than images because you could use image processing techniques to suppress the noise in the background and highlight the edges on the QR code. You could even look at autoencoder and CNN related feature extraction.

$\endgroup$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.