joobei / beatmup

Beatmup: image and signal processing library

Overview

Beatmup is an extensible parallel image and signal processing and rendering framework.

• It makes easy the execution of image processing tasks in parallel.
  • Image processing operations are implemented as asynchronous tasks. The tasks are submitted to the execution engine to perform the processing operations in asynchronous fashion.
  • Images might be processed by multiple workers in parallel for maximum speed.
• Beatmup enables hardware acceleration for imaging tasks through OpenGL and OpenGL ES in a transparent platform-independent way.
  • It allows for user-defined shaders for rendering and image processing
  • It uses GPU to infer neural networks in a vendor-agnostic fashion
• Beatmup natively supports 1, 2, 4, 8, 24 and 32-bit integer bitmaps and 32, 96 and 128-bit floating point bitmaps.
  • Enables 24-bit RGB bitmaps support in Android
  • Provides pixel arithmetics allowing to implement processing operations in a format-independent way
• It implements a number of predefined tasks providing a basic image processing toolkit.
  • A simple layered renderer
  • Basic color filters and color space conversion
  • Image resampling and pixel format conversion
    • x2, a neural network-based upscaling technique
  • Flood fill and contours extraction
• Beatmup runs in linux, Windows and Android. It works on Raspberry Pi and uses its GPU as well.
  • Android Java API is available.
    • Provides access to the camera. Enables realtime camera preview processing with custom shaders
  • Compiles with gcc, clang, MSVC
  • Few to no dependencies; git and CMake is all you need
• Beatmup evolves as fast as it can!
  • Basic OpenSL ES and AAudio support to play audio (Android)
  • Efficient audio signal graph plotting
  • Video decoding support (Android)
  • Neural networks inference using OpenGL (in progress)

Quick start

There is a bunch of test apps showcasing the use of Beatmup. To get things running, make sure you have Git and CMake installed, then open your favourite terminal and pick one of the copy-paste recipes that follow according to your OS.

Linux

To build X2 app upscaling an image using a neural net inferred with OpenGL you may proceed as follows:

git clone https://github.com/lnstadrum/beatmup.git
git submodule init
git submodule update
cd beatmup
mkdir build
cd build
cmake -DUSE_OPENGL=ON ../apps
make X2

Use -DUSE_GLX=ON instead -DUSE_OPENGL=ON if there are issues. On Raspberry Pi replace it with -DUSE_EGL=ON.

You can then feed the app with an image of your choice and get the upscaled result as follows:

./X2 <your-image>.bmp x2-result.bmp

Windows

To build the X2 app in Windows you may install Visual Studio (tested on 2013 Express and 2019 Community free editions) and generate the VS solution as follows:

git clone https://github.com/lnstadrum/beatmup.git
cd beatmup
git submodule init
git submodule update
mkdir build
cd build
cmake -DUSE_OPENGL=ON ../apps

Then open the .sln solution file in Visual Studio and build the executable.

Android

There is Android project containing a library and a test app modules (lib and app). To build them install Android Studio and Android NDK bundle, clone the repository and open the project situated in android folder in Android studio.

If for some reason you do not want to bother you with Android studio but you still want to try out the test app, you can build an apk in a docker container. If you have docker installed, the following script will build beatmupApp.apk installation package which you may then copy to your Android phone and open to install the app:

git clone https://github.com/lnstadrum/beatmup.git
git submodule init
git submodule update
cd beatmup
cp .github/workflows/make-apps/* .
docker build . -t android-build
docker run -v $(pwd):/android-build android-build