• Hardware Platform (Jetson / GPU) Jetson Orin Nano
• DeepStream Version 7.0
• JetPack Version (valid for Jetson only) 6.0
Hi, I’m trying to extract the position of objects detected with a probe. The PGIE has a batch-size=9, so I run through 9 frames per batch. When I extract the position of the objects, I get the right dimensions and coordinates, but it’s never the right frame.
I use batch_id to find my way through the frame count, it always goes from 0 to 8 but I don’t touch frame_meta.frame_num because it gives me the batch number since the start of the program. The OSD always shows the objects surrounded by a red rectangle in the right place, so I don’t understand why I can’t capture the right frame number.
For the record, I don’t use GLib.MainLoop() because I use a state machine on the side and it simplifies my work.
Here is my probe :
def tracker_src_pad_buffer_probe(self,pad,info,u_data):
gst_buffer = info.get_buffer()
batch_meta = pyds.gst_buffer_get_nvds_batch_meta(hash(gst_buffer))
l_frame = batch_meta.frame_meta_list
while l_frame is not None:
# 1 frame (1 tile)
try:
frame_meta = pyds.NvDsFrameMeta.cast(l_frame.data)
except StopIteration:
break
batch_id = frame_meta.batch_id
total_batch_number = frame_meta.frame_num # Batch number since the programm started
l_obj=frame_meta.obj_meta_list
while l_obj is not None:
try:
obj_meta = pyds.NvDsObjectMeta.cast(l_obj.data)
except StopIteration:
break
class_id = obj_meta.class_id
if self.run_calibration:
left = obj_meta.rect_params.left
top = obj_meta.rect_params.top
right = left + obj_meta.rect_params.width
bottom = top + obj_meta.rect_params.height
center = ((right+left)//2,(bottom+top)//2)
# Positions always correct but when I look at the batch_id, it never matchs.
try:
l_obj = l_obj.next
except StopIteration:
break
try:
l_frame = l_frame.next
except StopIteration:
break
return Gst.PadProbeReturn.OK
Do you have any ideas, or do you need more information? Thank you in advance.