zoukankan      html  css  js  c++  java
  • Ubuntu 18.04上DeepStream安装

    DeepStream主页:https://developer.nvidia.com/deepstream-sdk

    DeepStream Development Guide:https://docs.nvidia.com/metropolis/deepstream/4.0/dev-guide/

    DeepStream SDK API:https://docs.nvidia.com/metropolis/deepstream/4.0/dev-guide/DeepStream_Development_Guide/baggage/index.html

    环境搭建:

    1)要求

    • Ubuntu 18.04
    • Gstreamer 1.14.1
    • NVIDIA driver 418+
    • CUDA 10.1
    • TensorRT 5.1.5

    2)执行nvidia-smi查询Nvidia驱动版本

    Driver Version: 418.87.00

    执行cat /usr/local/cuda/version.txt查询CUDA版本

    CUDA Version 10.1.243

    执行cat /usr/local/cuda/include/cudnn.h | grep CUDNN_MAJOR -A 2查询cudnn版本

    7.6.3

    执行cat /proc/version查询ubuntu系统版本

    18.04.1

    Gstreamer安装:

    apt-get install libgstreamer1.0-0 gstreamer1.0-plugins-base gstreamer1.0-plugins-good gstreamer1.0-plugins-bad gstreamer1.0-plugins-ugly gstreamer1.0-libav gstreamer1.0-doc gstreamer1.0-tools gstreamer1.0-x gstreamer1.0-alsa gstreamer1.0-gl gstreamer1.0-gtk3 gstreamer1.0-qt5 gstreamer1.0-pulseaudio

    验证Gstreamer:dpkg -l | grep gstreamer

    TensorRT安装:参考https://docs.nvidia.com/deeplearning/sdk/tensorrt-install-guide/index.html#installing-tar

    下载:https://developer.nvidia.com/nvidia-tensorrt-5x-download,需要先登录NVIDIA账号

    tar xzvf TensorRT-5.1.5.0.Ubuntu-18.04.2.x86_64-gnu.cuda-10.1.cudnn7.5.tar.gz

    sudo vim ~/.bashrc,添加:

    source ~/.bashrc

    cd TensorRT-5.1.5.0/python/

    sudo -H pip3 install tensorrt-5.1.5.0-cp36-none-linux_x86_64.whl

    (若没安装pip,先安装pip3:sudo apt install python3-pip)

    cd ../uff   (计划将TensorRT与TensorFlow一起使用时,安装uff才是必要的)

    sudo -H pip3 install uff-0.6.3-py2.py3-none-any.whl

    which convert-to-uff

    若提示错误:

    先安装protobuf:

    下载:https://github.com/protocolbuffers/protobuf/releases

    cd protobuf-3.9.2/
    ./autogen.sh
    ./configure --prefix=/usr/local/protobuf
    sudo make -j12
    sudo make install
    sudo ldconfig
    sudo vi ~/.bashrc
    添加:PATH=/usr/local/protobuf/bin:$PATH
    source ~/.bashrc

    cd ../graphsurgeon

    sudo -H pip3 install graphsurgeon-0.4.1-py2.py3-none-any.whl

    验证TensorRT

    cd TensorRT-5.1.5.0/samples/sampleMNIST/

    make

    cd ../../bin

    ./sample_mnist,如下所示,则安装成功

    安装其他依赖:

    sudo apt install
         libssl1.0.0
         libgstrtspserver-1.0-0
         libjansson4

    3)安装librdkafka

    git clone https://github.com/edenhill/librdkafka.git

    cd librdkafka/
    git reset --hard 7101c2310341ab3f4675fc565f64f0967e135a6a
    ./configure
    make
    sudo make install
    sudo mkdir -p /opt/nvidia/deepstream/deepstream-4.0/lib
    sudo cp /usr/local/lib/librdkafka* /opt/nvidia/deepstream/deepstream-4.0/lib
    4)安装DeepStream SDK

    下载:https://developer.nvidia.com/deepstream-sdk

    tar -xpvf deepstream_sdk_v4.0_x86_64.tbz2

    cd deepstream_sdk_v4.0_x86_64

    sudo tar -xvpf binaries.tbz2 -C /

    sudo ./install.sh

  • 相关阅读:
    day44
    day38
    day37
    day36
    作业35
    day35
    作业34
    day34
    day33
    Windows API 第三篇
  • 原文地址:https://www.cnblogs.com/walker-lin/p/11528243.html
Copyright © 2011-2022 走看看