How to use csi camera in deepstream-4.0.1 yolov3?

Accelerated Computing Intelligent Video Analytics DeepStream SDK
1

Thanks in advanced. I update something in source1_csi_dec_infer_resnet_int8.txt and run: deepstream-app -c …/…/samples/configs/deepstream-app/source1_csi_dec_infer_resnet_int8.txt, it shows yolov3_tiny result. On tx2, I want to see yolov3 result, how to implement it?

2

Hi,
Please check

deepstream_sdk_v4.0.1_jetson\sources\objectDetector_Yolo

You may apply your customization to this config file:

deepstream_sdk_v4.0.1_jetson\sources\objectDetector_Yolo\deepstream_app_config_yoloV3_tiny.txt
1 Like
3

Thanks. I want to see yolov3 result, not yolov3-tiny result, with csi camera. I upgrade deepstream-4.0, but I can’t find folder deepstream_sdk_v4.0.1_jetson, when I reinstall, I see:

nvidia@nvidia-desktop:~/Downloads$ sudo apt-get install ./deepstream-4.0_4.0.1-1_arm64.deb
Reading package lists... Done
Building dependency tree       
Reading state information... Done
[b]Note, selecting 'deepstream-4.0' instead of './deepstream-4.0_4.0.1-1_arm64.deb'
deepstream-4.0 is already the newest version (4.0.1-1).[/b]
The following package was automatically installed and is no longer required:
  libllvm6.0
Use 'sudo apt autoremove' to remove it.
0 upgraded, 0 newly installed, 0 to remove and 22 not upgraded.
4

Hi,
There is a config file for yoloV3 also.

deepstream_sdk_v4.0.1_jetson\sources\objectDetector_Yolo\deepstream_app_config_yoloV3.txt

You can follow README to run this first.

5

Please pay attention to “yolov3” result with csi camera. :)

6

Hi,
You can use CSI camera( nvarguscamerasrc ) by applying blow change to [source0] in deepstream_app_config_yoloV3.txt

[source0]
enable=1
#Type - 1=CameraV4L2 2=URI 3=MultiURI 4=RTSP 5=CSI
type=5
camera-width=1280
camera-height=720
camera-fps-n=30
camera-fps-d=1

Guidance of setting source group:
https://docs.nvidia.com/metropolis/deepstream/dev-guide/index.html#page/DeepStream_Development_Guide%2Fdeepstream_app_config.3.2.html%23wwpID0E0QB0HA

7

There is no option: 5=CSI in my deepstream_app_config_yoloV3.txt, when I update type=5, it reports:

nvidia@nvidia-desktop:/opt/nvidia/deepstream/deepstream-4.0/sources/objectDetector_Yolo$ deepstream-app -c deepstream_app_config_yoloV3.txt

(deepstream-app:10031): GStreamer-CRITICAL **: 09:34:11.473: passed '0' as denominator for `GstFraction'
** ERROR: <create_camera_source_bin:152>: Failed to link 'src_elem' (video/x-raw(memory:NVMM), width=(int)[ 1, 2147483647 ], height=(int)[ 1, 2147483647 ], format=(string){ NV12 }, framerate=(fraction)[ 0/1, 120/1 ]) and 'src_cap_filter' (video/x-raw(memory:NVMM), format=(string)NV12, width=(int)0, height=(int)0)
** ERROR: <create_camera_source_bin:187>: create_camera_source_bin failed
** ERROR: <create_pipeline:1045>: create_pipeline failed
** ERROR: <main:632>: Failed to create pipeline
Quitting
App run failed

deepstream_app_config_yoloV3.txt (3.46 KB)

8

Hi,
The config file is not right. Please add camera-width, camera-height, camera-fps-n, camera-fps-d along with type=5.