parking slot detection

my graduation project is about parking slot detection, i implemented it on my pc and it works well based on yolo!

the main idea of the algorithm is lane detection then see if the area between lane is empty or there is an object in it.

and now i want to implement it on openMv cam, how can i detect objects with it?

Hi, we don’t support yolo CNNs yet. However, we’re about to release CNN support on the OpenMV Cam at the end of the month. You’ll be able to use a sliding window CNN approach. We have a method that does the sliding part for you. So, you just have to make a classifier.

Here’s the video demo of CNN support: Re-loadable CNN on the OpenMV Cam M7/H7 with CMSIS-NN - YouTube

(While I say we don’t support YOLO CNNs I mean that we don’t do any post processing to fix up the CNN output and etc. We’ll be working on that kinda of stuff soon. In the mean time, just training a very simple network. Like LENET-6 simple. With one class output to detect between a car and not a car).

The network has to be brought onto to the OpenMV Cam using these scripts here: https://github.com/openmv/openmv/tree/master/ml/cmsisnn