1 June 1996 Optimal recovery of signal parameters from a few samples: one- and two-dimensional applications
Author Affiliations +
The results of some previous work describing a method for fast and precise recovery of the parameters of a signal from a few samples are discussed and extended. The method introduces the representation of the signal in the space of its samples and, through it, it derives a procedure for computing a parameter as a piecewise linear quasi-invariant of the representative manifold of the signal. The method originated in an application from high energy physics where the localization and the amplitude of a pulse must be determined from its samples. The pulse shape was supposed to be known but real-time processing (tens of nanoseconds) and high precision were imposed; linear processing could not fit the demands. We succeeded in obtaining an order of magnitude better precision than the linear processing concomitantly keeping the computation compatible with the real-time processing demands. This problem is of larger interest in measuring systems. Next we extended the algorithms to images, namely, for precisely determining the movement of an object in a series of images, or for super-resolution.
Vasile V. Buzuloiu, Marius Malciu, Viorel George Popescu, Constantin Vertan, "Optimal recovery of signal parameters from a few samples: one- and two-dimensional applications," Optical Engineering 35(6), (1 June 1996). https://doi.org/10.1117/1.601123

Back to Top