A practical way to implement Claerbout's imaging principle is to use match filters (crosscorrelation of the shot and receiver wavefields). Thereby, for each shot position, a partial image is obtained by matching the source and the receiver wavefields along the time dimension. Then, the image is formed by stacking the partial images at each subsurface location.
We propose a different imaging condition that also satisfies Claerbout's imaging principle. It consists of deconvolving the receiver wavefield by the source wavefield in the shot position/time (xs,t) domain. This 2-D deconvolution imaging condition has the advantage of improving the image resolution while keeping the resulting image stable.
In this paper we show the advantages of the 2-D deconvolution over crosscorrelation and 1-D deconvolution imaging conditions. To do this, we use a synthetic model with five dipping reflectors. Using this model, we demonstrate the advantages of 2-D deconvolution not only for image resolution but also for estimating the correct moveout of reflectors in the angle-domain Sava and Guitton (2003).