«

rknn_toolkit_lite2在rk3588环境部署

DJ 发布于 阅读:161


系统环境:Ubuntu server 22.04(ARM64平台)

  1. 使用apt安装python3、pip3、gcc、opencv、numpy工具。(推荐提前更换apt源为国内源)

    sudo apt update
    sudo apt install -y python3 python3-dev python3-pip gcc
    sudo apt install -y python-is-python3
    sudo apt install -y python3-opencv
    sudo apt install -y python3-numpy
  2. 使用pip执行whl文件安装rknn_toolkit_lite2及相关工具包。(推荐提前更换pip源为国内源)

    whl文件在rknn-toolkit2-master\rknn_toolkit_lite2\packages目录下,根据安装的python版本,选择对应的whl文件。

    # Python 3.7
    pip3 install rknn_toolkit_lite2-1.x.y-cp37-cp37-linux_aarch64.whl
    # Python 3.8
    pip3 install rknn_toolkit_lite2-1.x.y-cp38-cp38-linux_aarch64.whl
    # Python 3.9
    pip3 install rknn_toolkit_lite2-1.x.y-cp39-cp39-linux_aarch64.whl
    # Python 3.10
    pip3 install rknn_toolkit_lite2-1.x.y-cp310-cp310-linux_aarch64.whl
    # Python 3.11
    pip3 install rknn_toolkit_lite2-1.x.y-cp311-cp311-linux_aarch64.whl
  3. 运行官方example测试环境是否部署成功。

    将rknn-toolkit2-master\rknn_toolkit_lite2\examples目录下文件夹拷贝到RK3588设备中,打开dynamic_shape或resnet18目录,执行:

    python test.py

    参考运行结果:

    --> Running model
    resnet18
    -----TOP 5-----
    [812] score:0.999676 class:"space shuttle"
    [404] score:0.000249 class:"airliner"
    [657] score:0.000014 class:"missile"
    [833] score:0.000009 class:"submarine, pigboat, sub, U-boat"
    [466] score:0.000009 class:"bullet train, bullet"
    
    done

apt换源参考:

pip换源参考: