In traditional film photography, the scene was projected onto a photo-sensitive material that was either directly viewed (after development) or used to create an analogue print. Within this framework, it was essential that the light was captured, as we wished to view it. As photography evolved, this has been slightly relaxed, first in the form of negative film and later Bayer filters and digital sensors, but, by and large, we still capture the image as we wish to view it, at least spatially. One of the key advantages of the computational photography framework is that we are not restricted to this ‘identity’ mapping and can consider other, more flexible, invertible mappings. This flexibility can lead to cheaper and more efficient devices as well as increased precision enabling new sensing possibilities.
My current research in this area is motivated by Lippmann photography, a Nobel prize winning technique to capture colour using interference. After fully modelling this historical process, we are currently in the process of bringing Gabriel Lippmann’s approach into the realm of computational photography. From a consumer photography viewpoint, this work can get us closer to Lippmann’s perfect photograph; that is, a photograph indistinguishable from looking through a window at the scene. Beyond this, the tools are not limited to the visible spectrum and applications extend way beyond consumer photography.
- G. Baechler, A. Scholefield, L. Baboulaz, M. Vetterli. Sampling and Exact Reconstruction of Pulses with Variable Width; IEEE Transactions on Signal Processing. vol. 65, no. 10, pp. 2629–2644, May 2017.
- A. Scholefield, P. L. Dragotti. Quadtree Structured Image Approximation for Denoising and Interpolation; IEEE Transactions on Image Processing. vol. 23, no. 3, pp. 1226–1239, March 2014.
- L. Jospin, G. Baechler, A. Scholefield. Embedded polarizing filters to separate diffuse and specular reflection; Asian Conference on Computer Vision (ACCV), 2018.
- A. Scholefield, B. Bejar Haro, M. Vetterli. Shape from bandwidth: The 2-D orthogonal projection case; IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP), 2017.
- A. Scholefield, P. L. Dragotti. Accurate image registration using approximate Strang-Fix and an application in super-resolution; Proceedings of the 22nd European Signal Processing Conference (EUSIPCO), 2014.
- A. Scholefield, P. L. Dragotti. Image restoration using a sparse quadtree decomposition representation; IEEE International Conference on Image Processing (ICIP), 2009.
- A. Scholefield, P. L. Dragotti. Quadtree structured restoration algorithms for piecewise polynomial images; IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP), 2009.