MXRT1020 Camera image acquisition and image classification

[Morphinn]

I am trying to create an application that classifies images from a video camera. The problem so far is that the source code that I have found (opencv cnn classifier and Shark) uses files and I need a filesystem. I have tried to use Zephyr, but it does not support filesystems for the MXRT1020. I have even fiddled around with its files  (here are some things I have tried) but I do not understand really well what files I should modify. I have tought of embedding a model in the source code but I am not sure how to do that either.

My question is if I am on the right track and if I am, how should I advance?

Thank you!

[gusarambula · Accepted solution]

Hello Morphinn,

There is an image classification lab of eIQ that performs image classification that may be useful.

https://community.nxp.com/t5/eIQ-Machine-Learning-Software/eIQ-Transfer-Learning-Lab-with-TensorFlow...

This lab uses an existing TensorFlow image classification model, and it’s a bit different from some of the other solutions you are exploring.

From your post on Reddit you are writing your own SDHC driver. I would recommend seeing the sdcard examples from the i.MXRT1020 SDK in case that the driver may be useful (it may need some changes to be used in Zephyr), there is even a FATFS example.

I hope that this information helps!

Regards,
Gustavo

[Morphinn]

Thank you so much for taking your time to read my post thoroughly! I really appreciate the time you spent answering.

This is a great answer and is really what I've needed! The lab says that it is intended for the 1050 and 1060, but hopefully the 1020 works as well. I'll give it a try for sure!

Also, thank you for the recommendation for the drivers. I'll look into that too!