# DaWelter_h264decoder **Repository Path**: umico/DaWelter_h264decoder ## Basic Information - **Project Name**: DaWelter_h264decoder - **Description**: mirror: DaWelter https://github.com/DaWelter/h264decoder - **Primary Language**: Unknown - **License**: Not specified - **Default Branch**: master - **Homepage**: None - **GVP Project**: No ## Statistics - **Stars**: 0 - **Forks**: 0 - **Created**: 2021-11-25 - **Last Updated**: 2022-01-18 ## Categories & Tags **Categories**: Uncategorized **Tags**: None ## README H264 Decoder Python Module ========================== The aim of this project is to provide a simple decoder for video captured by a Raspberry Pi camera. At the time of this writing I only need H264 decoding, since a H264 stream is what the RPi software delivers. Furthermore flexibility to incorporate the decoder in larger python programs in various ways is desirable. The code might also serve as example for libav and pybind11 usage. Examples -------- You can do something like this ```python import h264decoder import numpy as np f = open(thefile, 'rb') decoder = h264decoder.H264Decoder() while 1: data_in = f.read(1024) if not data_in: break framedatas = decoder.decode(data_in) for framedata in framedatas: (frame, w, h, ls) = framedata if frame is not None: #print('frame size %i bytes, w %i, h %i, linesize %i' % (len(frame), w, h, ls)) frame = np.frombuffer(frame, dtype=np.ubyte, count=len(frame)) frame = frame.reshape((h, ls//3, 3)) frame = frame[:,:w,:] # At this point `frame` references your usual height x width x rgb channels numpy array of unsigned bytes. ``` There are simple demo programs in the ```examples``` folder. ```display_frames.py``` is probably the one you want to take a look at. Requirements ------------ * Python 3 * cmake for building * libav / ffmpeg (swscale, avutil and avcodec) * pybind11 (will be automatically downloaded from github if not found) For the example scripts * matplotlib * numpy I tested it on * Ubuntu 18, gcc 9, Anaconda environment with Python 3.7, Libav from Ubuntu repo. * Windows 10, Visual Studio Community 2017, Anaconda environment with Python 3.7, FFMPEG from vcpkg. Building and Installing ----------------------- ### Windows The suggested way to obtain ffmpeg is through [vcpkg](https://github.com/microsoft/vcpkg). Assuming all the setup including VC integration has been done, we can install the x64 libraries with ```cmd vcpkg.exe install ffmpeg:x64-windows ``` We can build the extension module with setuptools almost normally. However cmake is used internally and we have to let it know the search paths to our libs. Hence the additional ```--cmake-args``` argument with the toolchain file as per vcpkg instructions. ```bash python setup.py build_ext --cmake-args="-DCMAKE_TOOLCHAIN_FILE=[path to vcpkg]/scripts/buildsystems/vcpkg.cmake" pip install -e . ``` The ```-e``` option installs symlinks to the build directory. Useful for development. Leave it out otherwise. ---------------------------------------------- Alternatively one can build the extension module manually with cmake. From the project directory: ```cmd mkdir [build dir name] cd [build dir name] cmake -DCMAKE_TOOLCHAIN_FILE=[path to vcpkg]/scripts/buildsystems/vcpkg.cmake -A x64 .. cmake --build . ``` ### Linux Should be a matter of installing the libav or ffmpeg libraries. On Debian or Ubuntu: ```bash sudo apt install libswscale-dev libavcodec-dev libavutil-dev ``` And then running ```bash pip install . ``` in the project directory. History ------- ### v2 For Python 3. Switch to PyBind11. Module renamed from libh264decoder to h264decoder! Support installation via setuptools. ### v1 For Python 2.7. Depends on Boost Python. Project/Build file generation with CMake. Credits ------- * [Michael Welter](https://github.com/DaWelter). Original author. * [Martin Valgur](https://github.com/valgur). Switch to pybind11, nice build configuration and more. License ------- The code is published under the Mozilla Public License v. 2.0.