Hi, I’m working on a project to detect moving person in a room. I found the frogeye2020 example interesting but I couldn’t find anything regarding the camera. Any information about it will be appreciated.
Hi, it was something Arducam asked us to get working. However, they never made anything but camera samples. So, it’s not a product.
Once the new version of the IDE is released this example will be hidden.
1 Like
Oh, thanks, no wonder I couldn’t find it anywhere…
Is there any other way to accomplish my goal with an OpenMV H7?
I’m current working on training a tensorflow model to recognize persons in a room, but I’d like to know if there are alternatives to it. H7 seems to be a little limiting in that area.
There’s literally a person detector model from TensorFlow: https://github.com/openmv/tensorflow-lib/tree/master/libtf/models
We don’t include it in the camera by default anymore though since it eats 300KB+ of flash space.
1 Like
Thank you for the respnose.
It looks like that model only detects if there is a person in the picture. I’m looking to track every pedestrian’s position in the view like this image below. Sorry if I didn’t make it clear.
The best you can do is to use edge impulse’s FOMO model. It’s like YOLO but it just gives centroids. Otherwise, it’s pretty good.
1 Like