Faults are critical to the accumulation of hydrocarbon and manifest themselves as abrupt, gradual or gentle
changes of seismic amplitude. However an important element of entrapment is the presence of numerous subtle
faults whose identification with computer–based algorithm poses a major challenge. Traditionally, edge
detection techniques such as coherence, semblance, Hilbert transform (HT), etc are employed in evaluating
faulted hydrocarbon prospects by examining trace to trace similarity in data and to unmask subtle events.
However, these have high sensitivity to noise and suffer from computational truncation, and are therefore
unreliable. This study focuses on the application of generalized Hilbert transform (GHT) of seismic amplitude
data in the interpretation of 3D seismic data from the Niger Delta. The GHT is a windowed conventional Hilbert
transform. It extends the HT by introducing a window and an order. HT is of order one, and one of the possible
orders of implementation of the GHT. The GHT is less sensitive to noise and gives better resolution of subtle
features than conventional edge detection techniques. The algorithm adopted is based on fast Fourier transform
techniques and was developed from basics and outside oil-industry interpretational platforms using standard
processing routines. Preliminary results of the algorithm, when implemented on both oil-industry and general
interpretational platforms gave convincing images. The generalized Hilbert transform of the thin bed reservoir
revealed subtle faults and provided an enhanced level of evaluating a prospect. This is capable of improving
reservoir production and performance.