Walsh Transforms in Seismic-Event Detection

Several approaches to detecting event signals in background noise are presented, and Walsh transforms are shown to be an efficient tool in implementing some of these. An experiment is reported in which four categories of detectors are represented, two using Walsh transforms in their implementation. The two Walsh-based detectors and one other have some capability of recognizing a spectral change in the data; these three outperformed a simple power detector in detecting low-level events. Of the three, the two Walsh-based ones have faster execution speed. Although these experimental results are for seismic signals and background, the usefulness of Walsh-transform-based detectors may extend to many other applications as well.