Highly Efficient and Data Compressed Ultrafast Single-Pixel Imaging based on Photonic Time-Stretch

The research presented in this thesis is focused on highly efficient and data compressed ultrafast single pixel imaging (SPI) systems based on photonic time stretch (PTS) technique. Three ultrafast SPI systems are presented and analysed with unique features of low-cost, compact, highly efficient and optical data compression.

Secondly, data compressed ultrafast photonic time stretch imaging is investigated with the help of a spatial mask for spatial domain compressed sensing. In practice, a spatial light modulator (SLM) is utilized as a passive optical random pattern modulator, namely, spatial mask, in spatial domain. This combines the benefit of compressed sensing (CS) and PTS techniques. And a high speed CS imaging system is obtained with a compression ratio of 55.6%. Besides, time-domain CS applied in ultrafast real-time optical coherent tomography (OCT) is experimentally demonstrated as well.

Finally, an all-optical CS imaging system based on PTS and multimode interference using a multimode fibre (MMF) is demonstrated. The MMF acts as a low-cost random optical speckle pattern generator based on ultrafast wavelength tuning in PTS. Each wavelength of the optical light generates a repeatable and stable random optical speckle pattern, which has the feature of low- correlated relation between different optical speckle patterns. This technique can overcome the speed limit in existing CS photonic time stretch imaging, where imaging speed is much lower than the pulse repetition rate.