Image inference in Deepstream Python

Accelerated Computing Intelligent Video Analytics DeepStream SDK
1

Please provide complete information as applicable to your setup.

• Hardware Platform (Jetson / GPU) Jetson
• DeepStream Version 6.3
• JetPack Version (valid for Jetson only) 5.1.2
• Issue Type( questions, new requirements, bugs) bugs

Hi,
I am trying to implement a Python DeepStream pipeline to infer a JPEG image with filesrc + jpegparser + nvv4l2decoder. You can see my pipeline below:

image
image4837×384 102 KB

However, when I run the pipeline, I encounter the following error. I am using a probe function at fakesink, but it is not being called. Could you tell me what the problem might be and how I can solve it?

Opening in BLOCKING MODE 
Opening in BLOCKING MODE 
0:00:00.369614733   119     0x137f3180 WARN                    v4l2 gstv4l2object.c:4561:gst_v4l2_object_probe_caps:<nvv4l2-decoder:src> Failed to probe pixel aspect ratio with VIDIOC_CROPCAP: Unknown error -1
0:00:00.369660399   119     0x137f3180 WARN                    v4l2 gstv4l2object.c:2420:gst_v4l2_object_add_interlace_mode:0x12eec470 Failed to determine interlace mode
0:00:00.369684400   119     0x137f3180 WARN                    v4l2 gstv4l2object.c:2420:gst_v4l2_object_add_interlace_mode:0x12eec470 Failed to determine interlace mode
0:00:00.369702417   119     0x137f3180 WARN                    v4l2 gstv4l2object.c:2420:gst_v4l2_object_add_interlace_mode:0x12eec470 Failed to determine interlace mode
0:00:00.369719793   119     0x137f3180 WARN                    v4l2 gstv4l2object.c:2420:gst_v4l2_object_add_interlace_mode:0x12eec470 Failed to determine interlace mode
0:00:03.760757795   119     0x137f3180 INFO                 nvinfer gstnvinfer.cpp:682:gst_nvinfer_logger:<pgie> NvDsInferContext[UID 1]: Info from NvDsInferContextImpl::deserializeEngineAndBackend() <nvdsinfer_context_impl.cpp:1988> [UID = 1]: deserialized trt engine from :/models/pgies/trafficcam/resnet18_trafficcamnet_pruned.etlt_b1_gpu0_int8.engine
INFO: [Implicit Engine Info]: layers num: 3
0   INPUT  kFLOAT input_1         3x544x960       
1   OUTPUT kFLOAT output_bbox/BiasAdd 16x34x60        
2   OUTPUT kFLOAT output_cov/Sigmoid 4x34x60         

0:00:03.937637762   119     0x137f3180 INFO                 nvinfer gstnvinfer.cpp:682:gst_nvinfer_logger:<pgie> NvDsInferContext[UID 1]: Info from NvDsInferContextImpl::generateBackendContext() <nvdsinfer_context_impl.cpp:2091> [UID = 1]: Use deserialized engine model: /models/pgies/trafficcam/resnet18_trafficcamnet_pruned.etlt_b1_gpu0_int8.engine
0:00:03.966136049   119     0x137f3180 INFO                 nvinfer gstnvinfer_impl.cpp:328:notifyLoadModelStatus:<pgie> [UID 1]: Load new model:./configs/pgies/trafficcam/trafficcam.txt sucessfully
0:00:03.966474750   119     0x137f3180 WARN                 basesrc gstbasesrc.c:3600:gst_base_src_start_complete:<file-source> pad not activated yet
NvMMLiteBlockCreate : Block : BlockType = 256 
[JPEG Decode] BeginSequence Display WidthxHeight 640x360
0:00:04.068971965   119     0x137e2d20 WARN                    v4l2 gstv4l2object.c:4561:gst_v4l2_object_probe_caps:<nvv4l2-decoder:src> Failed to probe pixel aspect ratio with VIDIOC_CROPCAP: Unknown error -1
0:00:04.069050720   119     0x137e2d20 WARN                    v4l2 gstv4l2object.c:2420:gst_v4l2_object_add_interlace_mode:0x12eec470 Failed to determine interlace mode
0:00:04.069095330   119     0x137e2d20 WARN                    v4l2 gstv4l2object.c:2420:gst_v4l2_object_add_interlace_mode:0x12eec470 Failed to determine interlace mode
0:00:04.069122947   119     0x137e2d20 WARN                    v4l2 gstv4l2object.c:2420:gst_v4l2_object_add_interlace_mode:0x12eec470 Failed to determine interlace mode
0:00:04.069146948   119     0x137e2d20 WARN                    v4l2 gstv4l2object.c:2420:gst_v4l2_object_add_interlace_mode:0x12eec470 Failed to determine interlace mode
0:00:04.069635767   119     0x137e2d20 WARN                GST_PADS gstpad.c:4231:gst_pad_peer_query:<nvv4l2-decoder:src> could not send sticky events
0:00:04.069893185   119     0x137e2d20 WARN            v4l2videodec gstv4l2videodec.c:2305:gst_v4l2_video_dec_decide_allocation:<nvv4l2-decoder> Duration invalid, not setting latency
0:00:04.072161081   119     0x137e2d20 WARN          v4l2bufferpool gstv4l2bufferpool.c:1114:gst_v4l2_buffer_pool_start:<nvv4l2-decoder:pool:src> Uncertain or not enough buffers, enabling copy threshold
0:00:04.077497447   119     0x2d9fbf00 WARN          v4l2bufferpool gstv4l2bufferpool.c:1565:gst_v4l2_buffer_pool_dqbuf:<nvv4l2-decoder:pool:src> Driver should never set v4l2_buffer.field to ANY
nvstreammux: Successfully handled EOS for source_id=0
3

Try building the following pipeline, I think it should work.

gst-launch-1.0 -e nvstreammux name=mux batch-size=1 width=1920 height=1080 ! nvinfer config-file-path=/opt/nvidia/deepstream/deepstream/sources/apps/sample_apps/deepstream-image-decode-test/dstest_image_decode_pgie_config.txt batch-size=1 ! nvvideoconvert ! nvdsosd ! nvvideoconvert ! nvjpegenc ! filesink location=out.jpg filesrc location=/opt/nvidia/deepstream/deepstream/samples/streams/sample_720p.jpg ! jpegparse ! nvv4l2decoder ! nvvideoconvert ! "video/x-raw(memory:NVMM), format=(string)NV12" ! mux.sink_0
4

Thank you for your reply. Based on your GStreamer pipeline, I realized some mistakes in my code and was able to correct them. This works well.

5

This topic was automatically closed 14 days after the last reply. New replies are no longer allowed.