Docs
{{ g.name }} {{ pg.name }}
Docs/SDKs/iQS-Streampipe: How to Change the Custom Model and Video Source

iQS-Streampipe: How to Change the Custom Model and Video Source

SDKStreamingCustom ModelVideo Source

This SDK guide covers editing config.json to change the model, video source and channel count for iQS-Streampipe — for running the pipeline on your own video or your own detection model.

Supported Versions

Choose the iQ-Studio version that matches your platform. To check your BSP version, run cat /etc/innodisk/BSP-version on the platform.

Your platformiQ-Studio version
BSP v2.5.0 or laterv0.0.11 or later
BSP earlier than v2.5.0v0.0.10
Ubuntu image x09v0.0.10

For v0.0.10, run git checkout v0.0.10 in the iQ-Studio directory before ./install.sh. To return to the latest version, run git checkout main.

How to Deploy

  1. 1
    At least one UVC camera
    • 1080p/30fps (1920x1080 pixels)
    • MJPEG compression format
    Due to limited bandwidth, do not use more than 3 UVC cameras at once.
  2. 2
    At least one MP4 video uses the H.264 codec.

How to modify the content displayed, please refer to the config.json file in the current directory.

  1. 1
    Navigate to the iqs-streampipe directory and check the file in your current directory.
    {{ cb1.fname }} bash
    iqs-streampipe directory tree
  2. 2
    Open the config.json file in the current directory

    To open and edit the config.json file. (default parameter)

    {{ cb2.fname }} bash
    config.json default contents
  3. 3
    Modify config.json. We have changed the video under id: 1 to coco_detect.mp4. You may replace coco_detect.mp4 with your own video file if desired.
    This config is not the default config in the Docker container.

    Enter the Relative Path into the config file.

    config.json with id 1 source changed to coco_detect.mp4
  4. 4
    After you finish modifying.

    save and close the config.json.

    {{ cb3.fname }} text
  5. 5
    Run the iqs-launcher
    You must add the --other "-c config.json" option to change the videos.
    {{ cb4.fname }} bash
  6. 6
    Change the display of ID 1 to the specified view.
    The demo GIF may take some time to load. If it does not appear immediately, please wait.
    Display of ID 1 changed to the specified view

How to use a UVC camera to capture video frames for real-time prediction.

The demo GIF may take some time to load. If it does not appear immediately, please wait.

When a UVC camera's USB plug is connected to the platform's USB port, the video prediction will automatically switch to and display the real-time prediction results from the UVC camera feed.

When multiple cameras are plugged in, the results will be replaced based on channels, starting with channel 0, then channel 1, and so on.

UVC camera real-time prediction
Due to limited bandwidth, using more than 3 UVC cameras will cause video lag. If you use more than 4 cameras, it will cause an unexpected bug.

Known Issue

The current Qualcomm codec driver may under certain conditions, cause the system to restart unexpectedly. If this occurs, re-run the process to continue.

You are here {{ railNum }}/ {{ railTotal }} {{ railPct }}