You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Hi :
Deepstream 6.4
dsl 0.29
cuda 12,2
driver 535.104.12
When I initialized the pipeline and tried to add a new source to it, the analysis and inference seemed to have stopped.
def run(self):
# Since we're not using args, we can Let DSL initialize GST on first call
while True:
retval = self._CreatePerSourceComponents('pipeline', 1, self.video_url, None)
if retval != DSL_RETURN_SUCCESS:
break
# New Primary GIE using the filespecs above, with infer interval
retval = dsl_infer_gie_primary_new('primary-gie', self.inferConfigFile, self.modelEngineFile, 0)
if retval != DSL_RETURN_SUCCESS:
break
# New IOU Tracker, setting operational width and hieght
retval = dsl_tracker_new('iou-tracker', self.iou_tracker_config_file, 480, 272)
if retval != DSL_RETURN_SUCCESS:
break
# New Tiler, setting width and height, use default cols/rows set by source count
retval = dsl_tiler_new('tiler', TILER_WIDTH, TILER_HEIGHT)
if retval != DSL_RETURN_SUCCESS:
break
retval = dsl_tiler_tiles_set('tiler', columns=2, rows=2)
if retval != DSL_RETURN_SUCCESS:
break
retval = dsl_pph_custom_new('custom-pph',client_handler=self._save_frame, client_data=None)
if retval != DSL_RETURN_SUCCESS:
break
retval = dsl_pph_custom_new('custom-pph2',client_handler=self._tiler_sink_pad_buffer_probe, client_data=None)
if retval != DSL_RETURN_SUCCESS:
break
# Add the ODE Pad Probe Hanlder to the Sink Pad of the Tiler
retval = dsl_tiler_pph_add('tiler', 'custom-pph', DSL_PAD_SINK)
if retval != DSL_RETURN_SUCCESS:
break
retval = dsl_tiler_pph_add('tiler', 'custom-pph2', DSL_PAD_SINK)
if retval != DSL_RETURN_SUCCESS:
break
retval = dsl_sink_fake_new('fake-sink')
if retval != DSL_RETURN_SUCCESS:
break
# Add all the components to our pipeline
retval = dsl_pipeline_new_component_add_many('pipeline',['source-1', 'primary-gie', 'iou-tracker', 'tiler', 'fake-sink', None])
if retval != DSL_RETURN_SUCCESS:
break
### IMPORTANT: we need to explicitely set the stream-muxer Batch properties, otherwise the Pipeline
# will use the current number of Sources when set to Playing, which would be 1 and too small
retval = dsl_pipeline_streammux_batch_properties_set('pipeline', self.MAX_SOURCE_COUNT, 40000)
# retval = dsl_pipeline_streammux_batch_properties_set('pipeline', 1, 40000)
if retval != DSL_RETURN_SUCCESS:
break
retval = dsl_pipeline_streammux_dimensions_set('pipeline', self.max_width, self.max_height)
if retval != DSL_RETURN_SUCCESS:
break
retval = dsl_pipeline_eos_listener_add('pipeline', self._eos_event_listener, None)
if retval != DSL_RETURN_SUCCESS:
break
# Play the pipeline
retval = dsl_pipeline_play('pipeline')
if retval != DSL_RETURN_SUCCESS:
break
dsl_main_loop_run()
retval = DSL_RETURN_SUCCESS
break
# Print out the final result
print(dsl_return_value_to_string(retval))
dsl_pipeline_delete_all()
dsl_component_delete_all()
@rjhowell44 I compared the examples and found the cause of the problem:
If I use this, it will result in the above error: retval=dsl_source_uri_new (components. source, uri, False, False, 4)
If I use this, it looks like it runs normally: retval=dsl_source_uri_new(components. source, uri, False, 0, 0)
Hi :
Deepstream 6.4
dsl 0.29
cuda 12,2
driver 535.104.12
When I initialized the pipeline and tried to add a new source to it, the analysis and inference seemed to have stopped.
When I initialized the pipeline and tried to add a new source to it, the analysis and inference seemed to have stopped. The log is as follows:
tiler like this
Do you have any ideas? Thank you
The text was updated successfully, but these errors were encountered: