|
Project Topic
|
In this Project, we design and implement radically novel 3D imaging devices, quantum plenoptic cameras, which exploit both momentum-position entanglement and photon-number correlations to enable the typical refocusing and ultra-fast, scanning-free, 3D imaging capability of plenoptic devices, but with dramatically enhanced performances: i) diffraction-limited resolution, unattainable in standard plenoptic cameras; ii) an unprecedented large depth of focus (DOF), even 10 times larger than in standard imaging at the given resolution; 3) ultra-low noise, aiming at sub-shot noise performances. Both photon pairs from a chaotic light source and entangled beams from parametric down conversion will be employed to build, respectively, a prototype and a demonstrator of quantum 3D imaging. Furthermore, quantum Fisher information-based imaging will be employed for pushing the resolution at the quantum limit, achieving QPI with 3D super-resolution. However, for the quantum advantages of the developed devices to be effective and appealing to end-users, two main challenges need to be tackled. First, the single-shot advantage of conventional plenoptic devices is currently lost in quantum imaging, where many frames need to be acquired for correlation measurements to give acceptable SNR: If implemented with commercially available sCMOS cameras (e.g., 50 fps at full resolution), quantum plenoptic imaging requires 10^3-10^4 frames, and a total acquisition time ranging from tens of seconds to a few minutes. Also, USB3 data transferring to the computer (taking from a few minutes to half an hour) is a bottleneck. Second, the elaboration of this large amount of data requires high-performance computing and extremely long times (tens of hours for reconstructing the correlation function and retrieving the 3D image, or refocusing a 2D image). The engineering partners of the consortium shall enable addressing these challenges, aiming at speeding-up acquisition and elaboration by several orders of magnitude, reaching 100Hz acquisition speed and elaboration times of a few minutes. Unique ultra-fast high-resolution single-photon detector arrays (100 kfps, 512x512) and high-performance low-level programming of ultra-fast electronics (e.g., GPU, FPGA), combined with state of the art machine-learning-inspired and compressive sensing algorithms, as well as quantum tomography techniques are the tools we shall employ toward reaching of this ambitious goal. Routes toward exploitation of the QPI devices will be considered. The Project addresses the QuantERA call areas 5, specifically quantum imaging. The overall aim of the Project is to start the translational research for moving our novel quantum 3D imaging technology from lab to end-users, ready for inspiring novel applications, scientific routes, and industrial products. In this perspective, we merge scientific research and engineering for optimizing the performances of the developed prototype and demonstrators in terms of resolution, DOF, noise, and, most challenging, acquisition and elaboration speed. Key elements are world-class single-photon sensor arrays, as well as methods and algorithms for data acquisition, elaboration and analysis inspired by machine learning, compressive sensing, and quantum tomography, combined with high-performance low-level programming of fast computing platforms. Based on the enormous scientific, industrial and societal potential of high-speed 3D imaging at high resolution and low noise, the results of the Project are expected to generate novel imaging and diagnostic tools, in many branches of science: quantum plenoptic microscopes and endoscopes for biophotonics and biomedical imaging, quantum space imaging devices, quantum 3D cameras for both security and industrial inspection applications. The research is thus expected to open new scientific and technological possibilities, and to play a transformational role in technology and society.
|
