Nvidia Of GPU The product development kit is called DeepStream, Is based on GStreamer Development . therefore , master DeepStream, Must be from GStreamer Starting with . Let's see how the goods are installed first ！

1 install GStreamer-1.0

The installation steps are as follows ：

sudo add-apt-repository universe
sudo add-apt-repository multiverse
sudo apt-get update
sudo apt-get install gstreamer1.0-tools gstreamer1.0-alsa \ gstreamer1.0-plugins-base gstreamer1.0-plugins-good \ gstreamer1.0-plugins-bad gstreamer1.0-plugins-ugly \ gstreamer1.0-libav
sudo apt-get install libgstreamer1.0-dev \ libgstreamer-plugins-base1.0-dev \ libgstreamer-plugins-good1.0-dev \ libgstreamer-plugins-bad1.0-dev

2 see GStreamer-1.0 edition

The order is as follows ：

gst-inspect-1.0 --version

3 GStreamer plug-in unit

GStreamer It is composed of plug-ins to form a media streaming pipeline , For details, please refer to 《GStreamer List of official introductory courses , Quickly help you find the examples you need 》. In this paper, some practical , Talk about what can be used in Nvidia Jetson A very practical plug-in on .

3.1 OpenMAX plug-in unit

OpenMax Is the open multimedia acceleration layer （ English ：Open Media Acceleration, Abbreviation for OpenMAX）, One that doesn't need Authorization 、 Cross platform software abstraction layer , With C Software interface implemented by language , Used to deal with Multimedia . It is from Khronos Group Proposed criteria , They also maintain , The goal is to create a unified interface , Speed up the processing of a large number of multimedia materials .

GStreamer Provides a set of tools based on OpenMAX Plug in for , The name is called gst-omx.

3.1.1 Video decoder

Video decoder	explain
omxh265dec	OpenMAX IL H.265 Video decoder
omxh264dec	OpenMAX IL H.264 Video decoder
omxmpeg4videodec	OpenMAX IL MPEG4 Video decoder
omxmpeg2videodec	OpenMAX IL MPEG2 Video decoder
omxvp8dec	OpenMAX IL VP8 Video decoder
omxvp9dec	OpenMAX IL VP9 video decoder

3.1.2 Video encoder

Video encoder	explain
omxh264enc	OpenMAX IL H.264/AVC video encoder
omxh265enc	OpenMAX IL H.265/AVC video encoder
omxvp8enc	OpenMAX IL VP8 video encoder (supported with NVIDIA Jetson TX2/TX2i and NVIDIA Jetson Nano; not supported with NVIDIA Jetson AGX Xavier)
omxvp9enc	OpenMAX IL VP9 video encoder (supported with Jetson TX2 and Jetson AGX Xavier; not supported with Jetson Nano)

3.1.3 sink

video Sink	explain
nvoverlaysink	OpenMAX IL videosink element

I wrote an example , Can get Nvidia Nona Upper CSI Video information of interface camera , use omxh264enc code , And then use omxh264dec decode , Finally, send the screen to display .

gst-launch-1.0  nvarguscamerasrc ! omxh264enc ! h264parse ! omxh264dec ! autovideosink

among nvarguscamerasrc( Camera data capture ) 、h264parse(h.264 Code parsing ) We'll continue later . If you are right about gst-launch-1.0 Not familiar with the pipeline configuration parameters behind it , Please refer to 《 Explain profound theories in simple language , Find out quickly gst-launch-1.0 PIPELINE-DESCRIPTION Pipeline description Syntax 》.

3.1.5 gst-omx No longer recommended

However ,Nvidia The official statement announced that ,Linux for Tegra (L4T) Release 32.1 It is not recommended to use gst-omx.

I searched it , I found an older version called gst-openmax. If that friend knows L4T Why not recommend it , You can leave me a message .

About Tegra
2008 year 2 month , Has already had a keen insight into the potential of the mobile Internet market NVIDIA It is very shocking to release a heavyweight cross-border product to the industry ——.
Tegra It is a system design using single chip microcomputer （system-on-a-chip） chip , It integrates ARM Architecture processor and NVIDIA Of GeforceGPU, And built-in other functions , Products are mainly for small equipment . and Intel With PC Starting from x86 Architecture compared to ,ARM Architecturally Tegra It's more like a development based on mobile phone processor . It doesn't work x86 PC Upper Windows XP Isooperating system , But it has been used on mobile phones for many years ARM The architecture of lightweight operating system can better meet its needs of high speed and low power consumption .
Brand promotion editor Broadcast
In order to better carry out NVIDIA（ Ying Wei Da ） Our brand promotion activities , send NVIDIA（ Ying Wei Da ） Products are more popular and widely remembered in China , from 2009 year 11 month 30 The date of ,NVIDIA（ Ying Wei Da ）Tegra Officially use Chinese names in China －－“ Toure ”.“NVIDIA（ Ying Wei Da ）Tegra（ Toure ）” Chinese and English combination name , Used in all NVIDIA（ Ying Wei Da ） pr 、 In sales and other promotional materials , as well as NVIDIA（ Ying Wei Da ） Partner's public relations 、 In sales and other promotional materials .“ Toure ” Must be connected with “Tegra” Combination of English names , It can't be used alone , And the order cannot be reversed , namely “Tegra” before ,“ Toure ” After .

---

For more information, please see the following article … …