I am currently working with the ZED 2i camera, both with and without object detection, and it is functioning as expected.
NOTE : I’M USING PYTHON AND CAN’T USE OTHER PROGRAMMING LANGUAGES
My objective now is to implement object tracking based on user-defined (x, y) coordinates. Specifically, I want to track a designated detected object while eliminating bounding boxes associated with all other objects.
However, I am facing two key challenges:
I am unsure how to retrieve real-time tracking information (including x, y, and z coordinates) directly, without first storing the data in a CSV file.
I need guidance on accessing ZED Metadata via GStreamer, particularly how to obtain the object array and other relevant information to facilitate object tracking.
I am not familiar with the python bindings of Gstreamer so I cannot provide an example, please contact the gstreamer support team for help on using the gstreamer API.
If Python is your only constraint, I can suggest using the ZED SDK’s python wrapper which provides more features and is more flexible than the GStreamer package.