arXiv:2508.03911v1 Announce Type: new
Abstract: Real-time ultrahigh-speed rendering of digital holograms from high-bitrate interferogram streams demands robust parallel computing and efficient data handling with minimal latency. We present Holovibes, a high-performance software engine that enables real-time holographic image reconstruction and short-time analysis at unprecedented throughput. Holovibes integrates spatial demodulation techniques, such as Fresnel transformations and angular spectrum propagation, with temporal analysis methods including short-time Fourier transform (STFT) and principal component analysis (PCA) in a unified pipeline. By leveraging CUDA-based GPU acceleration, multithreaded parallelism, and efficient buffering, the system achieves high-throughput, low-latency processing suitable for demanding computational imaging applications. We demonstrate sustained real-time hologram rendering of 256×256-pixel from interferograms acquired by a streaming camera at 71,400 frames per second on commodity hardware with no frame loss, while maintaining an end-to-end latency of 30 ms. The engine also supports simultaneous recording of raw or processed data, enabling high-speed acquisition workflows essential for experimental applications. This work represents a significant advance over prior digital holography systems and provides a versatile platform for ultra-high-speed, real-time computational imaging.