Dual-channel compressed ultrafast photography for Z-pinch dynamic imaging

Abstract

The compressed ultrafast photography (CUP) can capture non-repetitive time-evolving events at 7 × 1013 fps, which is anticipated to find a diverse range of applications in physics, biomedical imaging, and materials science. The feasibility of diagnosing ultrafast phenomenon of Z-pinch by using the CUP has been analyzed in this article. Specifically, a dual-channel CUP design has been adopted for acquiring high quality reconstructed images and the strategies of identical masks, uncorrelated masks, and complementary masks have been compared. Furthermore, the image of the first channel was rotated by 90° to balance the spatial resolution between the sweep direction and the non-sweep direction. Both five synthetic videos and two simulated Z-pinch videos were chosen as the ground truth to validate this approach. The average peak signal to noise ratio of the reconstruction results is 50.55 dB for the self-emission visible light video and 32.53 dB for the laser shadowgraph video with unrelated masks (rotated channel 1). The simulation results show that the time–space-evolving process of plasma distribution can be well retold, and the phenomenon of plasma instability can be accurately diagnosed by the dual-channel CUP with unrelated masks (rotated channel 1). This study may promote the practical applications of the CUP in the field of accelerator physics.

基于免训练自监督式神经网络算法实现压缩超快成像高保真图像重构研究

“…2014年, Gao等人 [12] 开发了一项名为压缩超快成像(compressed ultrafast photography, CUP)的技术 [13] , 通过将压缩感知 原理与时空偏转成像技术相结合, 实现了1000万帧/秒 的超快成像速度和超过300帧的超高成像帧数, 远超传 统单次多幅超快成像技术的性能. 随着近十年的发展, CUP凭借其优异的性能, 已经在观测光的反射和折 射 [12] 、飞秒脉冲时间聚焦 [14] 、皮秒光场时空演化 [15] 、 耗散光孤子 [ 1 6 ] 、激光烧蚀动力学 [ 1 7 ] 、超快光学弹 簧 [18] 、光学混沌 [19] 、电磁脉冲传播 [20] 、惯性约束聚 变 [21] 、荧光寿命成像 [22] 等领域得到应用. 然而, 由于 提出了一种结合遗传算法(genetic algorithm, GA)…”