[Jetson] How to set up it.

本身，它有自己的一套东西，才能发挥硬件的最大性能。

 

Overview

• Jetson AI Fundamentals - S3E1 - Hello AI World Setup

How to read the repo: jetson-inference

How to use the compiled command and pre-trained models to do inference.

 

Step by step

• Write Image to the microSD Card, on Windows.
• Remote control Jetson by Teamviewer

sudo dpkg --add-architecture armhf

sudo apt-get update

Download the latest teamviewer-host_armhf.deb from TeamViewer Host for Raspberry Pi

sudo dpkg -i teamviewer-host_armhf.deb

sudo apt-get -f install

Register Teamviewer account

sudo teamviewer setup

sudo systemctl enable teamviewerd.service

Next, login on the remote PC to control the Jetson board.

• Running the Docker Container

Launching the Container, download firslty, then ./docker/run.sh to enter into docker image.

dustynv/jetson-inference | r32.5.0 | ccc2a5f19dad | 6 weeks ago | 2.89GB

In the container, it connects to host by ./data.

• Install Gstreamer.
• Running Applications in docker environment:

./video-viewer /dev/video0, camera preview mode.

./detectnet /dev/video0, object detection demo.

• Camera Streaming and Multimedia

Send video data to host by RTP:

./video-viewer /dev/video0 rtp://<ip>:<port> , camera preview mode.

./detectnet /dev/video0 rtp://<ip>:<port>, object detection demo.

Receive video data from client:

gst-launch-1.0 -v udpsrc port=1234 caps = "application/x-rtp, media=(string)video, clock-rate=(int)90000, encoding-name=(string)H264, payload=(int)96" ! rtph264depay ! decodebin ! videoconvert ! autovideosink