The application of pattern recognition to oil and gas prospection is very recent, and results are still sparse. This paper presents an introduction to pattern recognition, a summary of previous applications in seismic processing, and several new pattern recognition approaches. Topics discussed include impluse recognition, horizon correlation, determination of the boundaries and nature of a reservoir, and two-dimensional Fourier (f-k) filtering to attenuate ground roll. To conclude, we examine the outlook for pattern recognition and propose increased use of the theory of fuzzy sets for pattern recognition applications in oil and gas exploration.
Download Full PDF Version (Non-Commercial Use)