LAN iChat （局域网内的视频聊天软件）

This project is a simple video chat tool.
It supports IPv4 and IPv6, using TCP protocal temporarily. Not finished yet , next step is to change into UDP. The current stable version is v0.2, the current developing version is v0.4. If you have thoughts for any of them, please e-mail me or just open your PR, I am very glad to learn from your brilliant idea.

注：授权实验楼发表的教程对应版本为 v0.2， 该部分代码在 v0.2/Linux 下。

New Features

v0.4 will provide the following abilities.

• Two hosts using different IP protocals can contact by a server, which has been provided in v0.3. However, v0.4 will reconstruct the classes (both server and client), the new version will use less bandwidth and memory.
• Everywhere, not only lan, can have an ichat using the server provided in v0.4. The server will be reconstructed using Golang 1.7.3 in v0.4, which can improve a lot in performance.
• One of the typical using: Your host is in a college lan, using IPv6; another host is in the wide Internet, like home or hotel, using IPv4. The two hosts can build a contact.

Platform (v0.2)

Windows

The script is written in python 3.5, using some packages, all listed below.

• cv2 : cv2 is the openCV packages for python. You can download openCV from its official site. Also, install by pip pip install opencv-python is needed. Then, install ffmpeg. Download it from here, I suggest you chose the binary file and just install it simply. After you install cv2, numpy and matplotlib will also be installed.
• pyaudio : packages for dealing with audio. You can install it by pip3 install pyaudio.
• Built-in packages: sys, struct, pickle, time, threading, argparse, zlib, and socket.

Linux

• To install OpenCV with Python support in Linux is a little troublesome. You can download opencv.sh to install it easily, or follow these steps.

$ mkdir ichat && cd ichat
$ sudo apt-get install build-essential libgtk2.0-dev libjpeg-dev libtiff5-dev libjasper-dev libopenexr-dev cmake python-dev python-numpy python-tk libtbb-dev libeigen2-dev yasm libfaac-dev libopencore-amrnb-dev libopencore-amrwb-dev libtheora-dev libvorbis-dev libxvidcore-dev libx264-dev libqt4-dev libqt4-opengl-dev sphinx-common texlive-latex-extra libv4l-dev libdc1394-22-dev libavcodec-dev libavformat-dev libswscale-dev
$ wget https://github.com/Itseez/opencv/archive/2.4.13.zip
$ unzip 2.4.13.zip
$ cd 2.4.13
$ mkdir release && cd release
$ cmake -D WITH_TBB=ON -D BUILD_NEW_PYTHON_SUPPORT=ON -D WITH_V4L=ON -D INSTALL_C_EXAMPLES=ON -D INSTALL_PYTHON_EXAMPLES=ON -D BUILD_EXAMPLES=ON -D WITH_QT=ON -D WITH_GTK=ON -D WITH_OPENGL=ON ..
$ sudo make
$ sudo make install
$ sudo nano /etc/ld.so.conf.d/opencv.conf   
$ Input /usr/local/lib and save
$ sudo ldconfig -v
$ sudo nano /etc/bash.bashrc
$ --Add the following two lines in the end of bash.bashrc
$ PKG_CONFIG_PATH=$PKG_CONFIG_PATH:/usr/local/lib/pkgconfig
  export PKG_CONFIG_PATH

• Attention, if you use Python3, you may need to specify the path. Add some options below.

$ cmake -D CMAKE_BUILD_TYPE=Release \
        -D CMAKE_INSTALL_PREFIX=/usr/local  \
           PYTHON3_EXECUTABLE=/usr/bin/python3 \
           PYTHON_INCLUDE_DIR=/usr/include/python3.4 \
           PYTHON_LIBRARY=/usr/lib/x86_64-linux-gnu/libpython3.4m.so \
           PYTHON3_NUMPY_INCLUDE_DIRS=/usr/local/lib/python3.4/dist-packages/numpy/core/include ..

Usage

Run python3 main.py <parameters>. Also, you can download the executable binary file ichat.exe from my cloud storage: here. Press Esc or Ctrl-C to stop this tool. The parameters are:

• --host: the remote host address you want to connect. Support IPv4 and IPv6. Default is 127.0.0.1.
• --port: the port you want to use. The remote should use the same port as yours. Default is 10087. The tool use 2 ports, and if you set port to be x, another port will be x+1.
• --version or -v: the IP version you want to use. Default is 4, if you want to use IPv6, remember setting --version=6.
• --level: the image level you want. It could be 0, 1, 2, 3. Values bigger than 3 is same to 3. Level 0 is the best quality, that means you will see the original image of the remote camera. However, it requires high network condition. Default is 1. If your network is blocked, please use level 2 or 3.
• --noself: The tool will display both the remote and your videos in default. If you do not need to show yourself on your screen, use --noself=True. Do not set any value to this flag if you need to observe your image.
• In Remote, press Esc to quit, in You, press Esc to close the window of yourself, but chat remains.
• One screenshot for usage is below. I had an ichat with myself. If you double click ichat.exe, you will see yourself.

Platform (v0.3)

Server

• v0.3 provides a file server.py, which is used to transmit data between IPv4 and IPv6. There are 4 instances created in running, the parameters are in format below. The first two parameters are the two ports used to receive connections from clients, and the next two parameters points the protocal they use. The last parameter shows the direction, True means client2 receives data from client1, False reverse it.

    contact1to2v = Remote_Server(start_port, start_port+5, 6, 4, True)
    contact2to1v = Remote_Server(start_port+1, start_port+4, 6, 4, False)
    contact1to2a = Remote_Server(start_port+2, start_port+7, 6, 4, True)
    contact2to1a = Remote_Server(start_port+3, start_port+6, 6, 4, False)

• Above is an example: Assume the start_port is 10000, then the server will listen 8 ports (10000 to 10007), the first 4 ports will connect with remote client A's audio server/client and video server/client, and the next 4 ports will connect with another remote client B.
• According to the configuration, client A is using IPv6, client B is using IPv4.
• As you can see, too much ports are used, so in the next version v0.4, I will reconstruct the models and use less ports. The current performance is not good, since many parameters are not adjusted. The next version v0.4 will fix most of the problems.

Client

• The data between client are all transmitted through server now. So in version v0.3, even two clients are in different LAN, they can still contact. However, the quality depends on the bandwidth of server. I want to find a way to let server make the two clients connect directly, however I have no idea until now.

Usage

In server, run python2 server.py to start server, remember to check the 4 instances. In client, run python3 main.py [OPTIONS] to start client. Press Esc to stop the client. The new version can assign the following options.

• --host: the remote server address. Support IPv4 and IPv6. Default is 127.0.0.1.
• --port: the server port, which is the start_port in server code.
• --version: the IP version you want to use. Default is 4, if you want to use IPv6, remember setting --version=6.
• --level: the image level you want. It could be 0, 1, 2, 3. Values bigger than 3 is same to 3. Level 0 is the best quality, that means you will see the original image of the remote camera. However, it requires high network condition. Default is 1. If your network is blocked, please use level 2 or 3.
• --noself: The tool will display both the remote and your videos in default. If you do not need to show yourself on your screen, use --noself=True. Do not set any value to this flag if you need to observe your image.

Update-logs

• 2016-9-23: Add this project, video is ok. Define v0.1.
• 2016-9-24: Add audio transmission, add quality settings, defile v0.2.
• 2016-10-6: Build repository.
• 2016-10-9: Add Linux version for v0.2.
• 2016-10-13: Add authorisation.
• 2016-11-1: Start updating to v0.3, supply a protocal-crossed version, two hosts using IPv4 and IPv6 can contact in this version.
• 2016-11-2: Finish v0.3, which is an unstable version, it will be improved a lot in v0.4.
• 2016-12-4: Build test, ready to update v0.4 before 2017.

License

All codes in this repository are licensed under the terms you may find in the file named "LICENSE" in this directory.

授权声明

我已授权实验楼使用此仓库中的代码并发表此项目教程，你可以在这里查看对应的教程。