Lecture: 'Compressive Sensing for the Photonic Mixer Device'

Compressive sensing (CS) is a novel mathematical theory that states that in most cases, real-world signals can be reconstructed from a number of measurements that is much lower than that suggested by classical sensing paradigms. Modern high resolution digital sensors, such as digital cameras, produce an enormous amount of data per acquisition and compression right after sensing becomes necessary. CS aims to perform compression at sensing, that is, carry out a reduced number of compressed measurements and afterwards recover the signal from them.

We aim to translate CS theory into high-impact applications, exploiting and integrating fundamental knowledge of the systems we deal with in order to gain a competitive advantage over naïve implementations. At the time being we have successfully applied CS to the Photonic Mixer Device (PMD), a core technology for 3D sensing, with encouraging results in terms of error reduction and resolution increase.

Currently we cooperate with the company pmdtec towards using CS to improve the capabilities of their future generation 3D cameras.