The reduction of scanning noise in raster scanned data.

It is common practice in many branches of observational science to build up a 2-dimensional image of an object by raster-scanning a sensor across the region of interest. In general, because of unavoidable instrumental offsets and drifts, the baselevel of each individual linear scan has to be corrected. A common technique is to subtract a linear baseline derived from the endpoints of each scan. The technique presented in this paper combines two or more data sets, taken with different scanning angles. The data are treated in the Fourier domain, providing an analytic, non-iterative and hence computationally efficient solution to the problem.