/rknn-3588-npu-yolo-accelerate

rknn-3588部署yolov5，利用线程池实现npu推理加速；Deploying YOLOv5 on RKNN-3588, utilizing a thread pool to achieve NPU inference acceleration.

Primary LanguageC++Apache License 2.0Apache-2.0

rk3588 Detect Accelerate

English | 简体中文

Project Overview

This project is an improved version of rknn-cpp-Multithreading, utilizing a thread pool to accelerate the processing and adding detailed comments to help beginners learn and use it more effectively.

Main Features

  • Thread Pool Acceleration: Uses thread pool technology to enhance model processing speed.
  • Educational Comments: Adds detailed comments to key parts of the code to facilitate understanding and learning for beginners.(The comments are in Chinese)
  • Open Source Foundation: Based on an open-source project, inheriting and extending its functionality.

Instructions

To successfully build and run this project, you need to meet the following requirements:

  • System Dependencies: OpenCV must be installed on your system.
  • Build Tools: Use CMake to build the project.

Model Information

This project uses the official model and converts it using the following tools:

  • Model Conversion Tool: Uses the official rknn-toolkit2 for model conversion.

Getting Started

  1. Ensure OpenCV and CMake are installed.
  2. Clone the repository to your local machine: 
    git clone https://github.com/wzxzhuxi/rknn-3588-npu-yolo-accelerate
  3. Navigate to the project directory and create a build directory: 
    cd rknn-3588-npu-yolo-accelerate-master
mkdir build && cd build
  4. Build the project using CMake: 
    cmake ..
make
  5. Running the project: First, return to the main directory 
    cd ..
Then run the following command:
```bash
./yolov5_thread_pool model video_source num_threads
Or run the shell script:
```bash
./yolorun.sh