mjpg-streamer-pplite

这是http://sourceforge.net/projects/mjpg-streamer/ 的分支，并通过input_raspicam插件增加了对Raspberry Pi相机的支持。

此版本嵌入了百度飞桨轻量级深度学习推理框架Paddle Lite，配合OpenCV 3.2，实现了深度学习模型推理视频流的结果通过mjpg-streamer得以在web浏览器上显示。

软硬件测试环境

本人使用的设备：树莓派3B

操作系统 ubuntu-18.04.4-preinstalled-server-arm64+raspi3 （64位系统）

此OS下载链接: https://pan.baidu.com/s/1unTK_lp1XHlVo1U7jYHCsQ 提取码: chp1

OpenCV版本 3.2.0

CMake版本 3.10.2

Paddle Lite版本2.8

环境配置

安装如下依赖：

sudo apt-get install cmake libjpeg8-dev libopencv-dev libprotobuf-c-dev libsdl-dev python3-dev python3-opencv

使用opencv输入还需要安装g++

sudo apt-get install gcc g++

运行程序

克隆项目至树莓派上，

然后执行脚本download_pplite2.8_libs.sh安装Paddle Lite 2.8的预测库。

git clone https://github.com/hang245141253/mjpg-streamer-pplite.git

cd mjpg-streamer-pplite/mjpg-streamer-experimental/

sh download_pplite2.8_libs.sh

make后执行start.sh（注意提前插入摄像头）

make && sh start.sh

然后通过局域网设备使用浏览器访问你树莓派ip的8080端口，即可查看到一个实时口罩检测的画面。

Discussion / Questions / Help

Probably best in this thread http://www.raspberrypi.org/phpBB3/viewtopic.php?f=43&t=45178

Authors

mjpg-streamer was originally created by Tom Stöveken, and has received improvements from many collaborators since then.

License

mjpg-streamer is free software; you can redistribute it and/or modify it under the terms of the GNU General Public License as published by the Free Software Foundation; version 2 of the License.

This program is distributed in the hope that it will be useful, but WITHOUT ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License for more details.

FAQ

-- The following OPTIONAL packages have not been found:

• PythonLibs
• Numpy
• SDL

CMake Error: The following variables are used in this project, but they are set to NOTFOUND.

sudo apt install python3-dev