WP1 – Imaging with Correlation
Work Package 1 – Imaging with Correlation
How many photons are really required to obtain an image? How can we image invisible gases? How can a camera see through smoke?
Research and development in Work Package 1 – Imaging with Correlation, exploits the quantum-inspired correlation between image and source to answer these questions through the development of single pixel cameras and ghost imaging systems coupled with advanced image reconstruction and theoretical development.
QuantIC’s Single Pixel Camera (SPC) has demonstrated its ability to detect images using only a single photon of light and have been able, for instance, to image leaking methane gas. Within the the first year of the Hub, our researchers have also transformed the SPC demonstrator from a suitcase-sized device to a demonstrator that sits in the palm of one’s hand, and this has attracted considerable commercial interest. Industry-led projects are ongoing under the QuantIC Partnership Resource Fund to develop a gas imaging camera, “Gas-Sight”, (Edgar,M et al, Scientific Reports, 5, 10669 (2015)) in collaboration with M Squared Lasers and a 3D ranging system proof of concept (Sun, M-J et al, Nature Comms, 7, 12010 (2016)) in collaboration with Leonardo.
Our second demonstrator, the wavelength transformation camera (Sun, M-J et al Optics Express 24(10),10476-10485 (2016), has been developed using a ghost imaging system where the sample is illuminated in the infrared, yet the image is recorded using a high-performance visible camera. The imaging system has been embodied into a microscope and is at a performance level where it can be demonstrated to potential users.
The data analysis and image reconstruction theoretical efforts in Work Package 1 have played a major role in supporting the improvement of the performance of our cameras. Our researchers have used both Gaussian and Poissonian noise models and have accelerated the implementation of these algorithms on a Graphics Processing Unit (GPU). Another development is a quantum correlation-based technique to reconstruct an image that has been corrupted by noise (Sonnleitner, M et al, Optica 2 (11), 950-957 (2015) and the practicality of this system is being investigated further in real environments.