The position saved when calculating the objects in the current node detect the 3d position of the object using global coordinates. We dont have a system that enables the detection of the objects and persist a tracking ID. For this I recommend to use a combination of Setting a limit of the objects we want to detect and create some type of logic based on that constraint and Kalman Filters (Take as inspiration this library https://github.com/wmuron/motpy).
The position saved when calculating the objects in the current node detect the 3d position of the object using global coordinates. We dont have a system that enables the detection of the objects and persist a tracking ID. For this I recommend to use a combination of Setting a limit of the objects we want to detect and create some type of logic based on that constraint and Kalman Filters (Take as inspiration this library https://github.com/wmuron/motpy).