全息技术是颗粒场诊断的重要技术，其在诊断颗粒三维空间位置、粒度及形状等方面具有难以替代的优势。随着数字全息重建算法、深度学习、高分辨率相机及短脉冲大能量激光器的成熟，数字全息与单沟槽微喷试验的具体需求逐渐契合。我们针对单沟槽微喷实验，在小于1ns的时间尺度上获取了1亿像素级的数字全息图像，随后基于神经网络算法辅助了颗粒目标的识别与重建，最终获取了微喷前端的结构分布及破碎颗粒的粒度分布。相比于传统光学全息，数字全息在图像信噪比上表现出更高的优势，有助于颗粒信息的精确提取，后续可能针对不同加载状态开展系列试验，获取更加丰富的试验数据。
For the irreplaceable superiority of revealing 3D position, size and shape of each single particle, holography is important for diagnosing particle field. The requirement of applying in-line digital holography in ejecta experiment is satisfied with the growing of holographic reconstruction algorithm, deep learning, high resolution camera and high-energy sub-nanosecond laser. We applied the sub-nanosecond single shot in-line digital holography for the ejecta particle field of high explosive loaded shock physics experiment, and got a digital hologram with the resolution more than a hundred million pixels. A neural network based algorithm is designed to help object recognition and refocusing. As a result, the structural and size distribution of ejecta field is reported. Compared to traditional optical holography, digital holography get rid of the negative effect of quadratic noise pollution can obtain higher signal-to-noise ratio, which is helpful for accurate extraction of particle information. In the future, a series of experiments may be carried out for different loading states to obtain richer experimental data.
 

ejecta,holograph,particle,Particle Size