rknn3588的用法

JobDocLS2026-03-17 8:12

环境搭建

安装rknn-toolkit2

python 复制代码

git clone https://github.com/rockchip-linux/rknn-toolkit2.git
cd rknn-toolkit2/packages/x86_64
conda create -n toolkit2 python=3.8
conda activate toolkit2
pip install -r requirement_cpp38.txt
pip install rknn_toolkit2-2.3.2-cp38-cp38-manylinux_2_17_x86_64.manylinux2014_x86_64.whl

验证是否安装成功
在 Python 环境中导入 RKNN 模块，检查是否成功：

python 复制代码

python3 -c "from rknn.api import RKNN; print('RKNN-Toolkit2 导入成功！')"

运行官方提供的测试脚本，验证完整功能：

python 复制代码

cd rknn-toolkit2/examples/onnx/yolov5
python3 test.py

模型转换

onnx转rknn
参考代码:
- https://github.com/airockchip/rknn_model_zoo/tree/main/examples/yolov8
- python convert.py .../model/yolov8n.onnx rk3588

c++推理

编译和构建
- https://github.com/airockchip/rknn_model_zoo/blob/main/docs/Compilation_Environment_Setup_Guide.md#linux-platform

python 复制代码

 1、下载交叉编译工具链
 https://releases.linaro.org/components/toolchain/binaries/6.3-2017.05/aarch64-linux-gnu/gcc-linaro-6.3.1-2017.05-x86_64_aarch64-linux-gnu.tar.xz
 2、解压文件
 tar Jxf gcc-linaro-6.3.1-2017.05-x86_64_aarch64-linux-gnu.tar.xz
 3、export GCC_COMPILER=~/gcc-linaro-6.3.1-2017.05-x86_64_aarch64-linux-gnu/bin/aarch64-linux-gnu
 4、./build-linux.sh -t rk3588 -a aarch64 -d yolov8

python代码推理

python 复制代码

cd python
# Inference with PyTorch model or ONNX model
python yolov8.py --model_path <pt_model/onnx_model> --img_show

# Inference with RKNN model
python yolov8.py --model_path <rknn_model> --target rk3588 --img_show