Time encoding and decoding of multidimensional signals with finite rate of innovation

In our previous work [1], we have developed a mathematical framework for time encoding and reconstruction of non-bandlimited signals, such as groups of 1D Diracs, pulses and piecewise constant signals. In this paper, we extend the principles of time encoding to the multidimensional case. We investigate a novel architecture for time encoding and perfect decoding of certain classes of non-bandlimited, which is based on diversity of pixels and which allows us to estimate the video sequence efficiently and accurately, using a small number of samples at each pixel.