此文档面向零基础用户详细介绍如何快速在计算机上使用 RKNN-Toolkit2 完成模型转换，并通过 RKNPU2 部署到 Rockchip 开发板上。本文所用示例已集成到 RKNN Model Zoo 中。 支持的平台：RK3562、RK3566系列、RK3568系列、RK3576系列、RK3588系列 开发板：BPI-W3, BPI-M7,BPI-AIM7, BPI-M5 Pro, BPI-CM5 Pro,...
RK3562/ RK3566/ RK3568/ RK3588/ RK3562/ RV1103/ RV1106 请使用 RKNN_toolkit_2 版本 六、推理检测 1. orangepi 板端进行图片推理检测 :~/rknpu2_1.5.2_20230825/examples/rknn_yolov5_demo/install/rknn_yolov5_demo_Linux$ ./rknn_yolov5_demo ./yolov5s_relu.rknn ./model/mans.jpeg AI检测代...
BR2_PACKAGE_RKNPU2用于启用或禁用rknpu2包,它提供了运行时库和服务,支持RK处理器中的神经网络加速功能。 BR2_PACKAGE_RKNPU2_ARCH用于指定rknpu2的架构,对于RK3588配置项BR2_ARCH=aarch64,因此BR2_PACKAGE_RKNPU2_ARCH=aarch64。 BR2_PACKAGE_RKNPU2_SOC指定rknpu2的目标系统,对于RK3588配置项BR2_PACKAGE_R...
# 进入 rknn_model_zoo/examples/yolov8/python 目录cd Projects/rknn_model_zoo/examples/yolov8/python# 运行 convert.py 脚本，将原始的 ONNX 模型转成 RKNN 模型# 用法: python convert.py model_path [rk3566|rk3576|rk3588|rk3562] [i8/fp] [output_path]python convert.py ../model/yolov8n.onnx...
支持RK3562, RK3566, RK3568, RK3588, RK3576 平台。 部分支持RV1103, RV1106 支持RK1808, RV1109, RV1126 平台。依赖库安装RKNN Model Zoo依赖 RKNN-Toolkit2 进行模型转换, 编译安卓demo时需要安卓编译工具链, 编译Linux demo时需要Linux编译工具链。这些依赖的安装请参考 https://github.com/airockchip/...
(yolo8) orangepi@orangepi5plus:~/230/rknn_model_zoo/examples/yolov8/python$ python yolov8.py --model yolov8n.rknn --img_save I rknn-toolkit2 version: 2.3.0 --> Init runtime environment I target set by user is: rk3588 E RKNN: [14:43:28.161] 6, 1 ...
支持RK3562, RK3566, RK3568, RK3588, RK3576 平台。 部分支持RV1103, RV1106 支持RK1808, RV1109, RV1126 平台。依赖库安装RKNN Model Zoo依赖 RKNN-Toolkit2 进行模型转换, 编译安卓demo时需要安卓编译工具链, 编译Linux demo时需要Linux编译工具链。这些依赖的安装请参考 https://github.com/airockchip/...
RKNN Model Zoois developed based on the RKNPU SDK toolchain and provides deployment examples for current mainstream algorithms. Include the process ofexporting the RKNN modeland usingPython APIandCAPIto infer the RKNN model. SupportRK3562,RK3566,RK3568,RK3588platforms. (RV1103,RV1106platforms support...
RKNN Model Zoois developed based on the RKNPU SDK toolchain and provides deployment examples for current mainstream algorithms. Include the process ofexporting the RKNN modeland usingPython APIandCAPIto infer the RKNN model. SupportRK3562,RK3566,RK3568,RK3588platforms. (RV1103,RV1106platforms support...
接着进入rknn_model_zoo目录下的对应python路径. cd /home/cxf/Projects/rknn_model_zoo/examples/yolov8_seg/python 执行下面指令将yolov8_seg.onnx模型转换成yolov8_seg.rknn模型. # 这里的 rk3588 是板子设备的型号. python convert.py train_seg_best.onnx rk3588 i8 train_seg_best.rknn ...