detectnet camera objects separate.

hello dusty, I’m sorry to ask you so often.
i running detectnet-camera
When objects overlap, boxes merge.
can separate objects?

Hi,

You can update post-processing here based on your requirement:
https://github.com/dusty-nv/jetson-inference/blob/master/c/detectNet.cpp#L692

Thanks.

I know where it is.
But I don’t know how to fix it.
my system will only operate if the object is separate
i have caffemodel.
clusterDetections() (caffe) need to be modified.
just one object by one bounding box…
do you have any examples? :(

Hi,

Sorry for the late update.

Suppose you can design an algorithm for your requirement.
For example, apply the NMS on the bounding box.
Just check all the output bounding box, and only restore the boxes with larger area.

Thanks.