A MATLAB toolbox from Cosys-Lab for high-performance acoustic imaging, featuring GPU & MEX-accelerated DMAS-CF beamforming and signal processing.
Find the latest release on Github or the MathWorks File Exchange for the toolbox. You can also clone or download this repository and use it from source.
Note that if you want to use any of the accelerated GPU and CPU-based MEX implementations for beamforming you will first need to compile these MEX files.
Once you have installed the toolbox you run the following command to automatically compile the source MEX files. Note that the GPU implementation requires an NVIDIA GPU!
enableGPUCompile = true;
clait.compileClaitMexFunctions(enableGPUCompile)This toolbox requires the following other toolboxes installed as well:
For more information on the implemented delay-multiply-and-sum (DMAS) beamforming technique, see our open-access publication here.
If you do use this toolbox to generate acoustic images, please concider citing our work as:
@ARTICLE{11363474,
author={Jansen, Wouter and Daems, Walter and Steckel, Jan},
journal={IEEE Access},
title={Delay-Multiply-And-Sum Beamforming for Real-Time In-Air Acoustic Imaging},
year={2026},
doi={10.1109/ACCESS.2026.3657901}}
Two example are available, to quickly open these files after installing the toolbox, run the following commands or find them manually in the examples folder.
clait.openClaitAcousticImageExample
clait.openClaitPlottingPolarEnergyscapeExampleThe toolbox has 8 functions: 2 main ones and 6 helper ones. The following lists all the available functions:
calculateAcousticImage:Primary high-level function. Executes the full imaging pipeline: Matched Filtering, Delay Calculation, Beamforming, and Post-processing. mlx).plotPolarEnergyscape:Primary high-level function. It plots advanced 2D polar coordinate visualizations of acoustic imaging data from CLAIT acoustic image generation function algorithms. mlx).calculateDMASCF:Core beamforming function. Computes the image using D(M)AS(-CF).calculateDelayMatrix:Calculates the necessary sample delays array steering to specified directions.generalizedMatchedFilter:Applies various matched filter and generalized correlation transforms (Normal, PHAT, ROTH, SCOT).fm_sweep:Helper function that enerates a hyperbolic FM sweep signal with Hanning windowing.normLog:Helper function that normalizes a matrix and converts it to a logarithmic (dB) scale.simulateMicrophoneData:Helper function that generates simulated microphone data from point targets in a scene.
All functions in this Toolbox exist within the clait namespace. For calling these functions you therefore have to add the namespace name to the beginning.
output = clait.functionName(...)
You will have to compile the MEX files after installing the Toolbox or when using this repository from source.
For compatibility C MEX compiler dependent on your operating system, see this resource: https://www.mathworks.com/support/requirements/supported-compilers.html
Make sure to correctly configure Matlab first for MEX compiling. Run the following command and make sure you either have the right compiler selected on Windows.
mex -setup c++Once you have installed the dependencies you can run the following command to automatically compile the source MEX files.
enableGPUCompile = true;
clait.compileClaitMexFunctions(enableGPUCompile)Go in your Matlab to the folder containing calculateDMASCFMex.cpp (found in the toolbox\+clait folder).
Use the command below to compile the function:
mex calculateDMASCFMexCPU.cppGo in your Matlab to the folder containing calculateDMASCFGPU.cu (found in the toolbox\+clait folder).
Use the command below to compile the function:
mexcuda calculateDMASCFMexGPU.cuYou can use the -v argument to see the compilation process. If you run into compatibility errors you can also try the following command to get passed them.
mexcuda NVCCFLAGS="--allow-unsupported-compiler" -D_ALLOW_COMPILER_AND_STL_VERSION_MISMATCH -v calculateDMASCFGPU.cuThis project is released under the CC-BY-NC-SA-4.0 license.