Morgan Zak, Cho Youngjun, Subramanian Sriram
Department of Computer Science, University College London, London, UK.
Commun Eng. 2025 Jun 5;4(1):101. doi: 10.1038/s44172-025-00439-w.
Ultrasound holography, pivotal in applications like mid-air haptics, volumetric displays and 3D printing, faces challenges in the crafting and timely measurement of precise acoustic holograms. Current methods, bench-marked via simulations due to slow measurement times, often neglect real-world complexities such as non-linearity and hardware tolerances, leading to discrepancies between predicted and observed results. Here we introduce a real-time 2D thermographic measurement technique orders of magnitude faster than microphone scans, although with reduced accuracy and no phase information, with a maximum peak pressure of 4.25 kPa validated and a demonstrated average accuracy of 2.5% in peak measurement. Higher pressures of approximately 12 kpa were captured, but validation was limited by the microphone. This method is grounded in thermo-viscous acoustic models for thin-ducts and micro-perforated plates. Finally, we integrate this with holography algorithms to propose a camera-in-the-loop algorithm that employs real-time measurement, enabling targeted data acquisition and on-line training of acoustic holography algorithms. This method achieved a 1.7% error in pressure with a single point compared to 7.8% for a conventional algorithm, and a 3.6% error and 4.2% standard deviation for 16 points compared to 9.7% and 6.9%. We further envisage this method as being capable of measuring acoustic streaming.
超声全息术在诸如空中触觉、立体显示和3D打印等应用中起着关键作用，但在精确声学全息图的制作和实时测量方面面临挑战。由于测量时间较慢，当前方法通过模拟进行基准测试，往往忽略了诸如非线性和硬件公差等现实世界的复杂性，导致预测结果与观测结果之间存在差异。在此，我们引入一种实时二维热成像测量技术，其速度比麦克风扫描快几个数量级，尽管精度有所降低且没有相位信息，验证的最大峰值压力为4.25kPa，在峰值测量中显示的平均精度为2.5%。捕获了约12kPa的更高压力，但验证受到麦克风的限制。该方法基于薄管道和微穿孔板的热粘性声学模型。最后，我们将其与全息算法集成，提出一种循环相机算法，该算法采用实时测量，能够进行有针对性的数据采集和声学全息算法的在线训练。与传统算法相比，该方法在单点压力测量中的误差为1.7%，而传统算法为7.8%；对于16个点，该方法的误差为3.6%，标准偏差为4.2%，而传统算法分别为9.7%和6.9%。我们进一步设想该方法能够测量声流。
