1# Yolo-v5 demo 2 3# 导出rknn模型 4 51. 进到examples/RV1106_RV1103/rknn_yolov5_demo/convert_rknn_demo/yolov5目录下，执行如下命令，可生成rknn模型: 6 7```sh 8 cd examples/RV1106_RV1103/rknn_yolov5_demo/convert_rknn_demo/yolov5 9 python onnx2rknn.py 10``` 11 12## arm Linux Demo 13 14### 编译 15 16RV1106/RV1103编译脚本均为 `build-linux_RV1106.sh`，设置交叉编译器所在目录的路径 `RK_RV1106_TOOLCHAIN`，例如修改成 17 18```sh 19export RK_RV1106_TOOLCHAIN=~/opts/toolchain/arm-rockchip830-linux-uclibcgnueabihf/bin/arm-rockchip830-linux-uclibcgnueabihf 20``` 21 22然后执行： 23 24```sh 25./build-linux_RV1106.sh 26``` 27 28### 推送执行文件到板子 29 30连接板子的usb口到PC,将整个demo目录到 `/userdata`: 31 32```sh 33adb push install/rknn_yolov5_demo_Linux /userdata/ 34``` 35 36### 运行 37 38```sh 39adb shell 40cd /userdata/rknn_yolov5_demo_Linux/ 41 42export LD_LIBRARY_PATH=/userdata/rknn_yolov5_demo_Linux/lib 43./rknn_yolov5_demo model/RV1106/yolov5s-640-640.rknn model/bus.jpg 44``` 45 46Note: LD_LIBRARY_PATH 必须采用全路径 47
