Semi-Automated Detection of 4D Anomalies
Paul Meldahl, Nayeff Najjar, T. Oldenziel-Dijkstra, H. Ligtenberg · 2002
P315 SEMI-AUTOMATED DETECTION OF 4D ANOMALIES P. MELDAHL 1 N. AL-NAJJAR 1 T. OLDENZIEL 2 and H. LIGTENBERG 2 Abstract 1 In this paper a simple yet sophisticated time-lapse object detection procedure is presented and illustrated on a North Sea field. Using multiple directive attributes and neural networks the visualisation of time-lapse anomalies is enhanced considerably. The procedure involves two parts; an analysis phase and the actual 4D-detection phase. In the analysis phase both single and multi-attribute analyses are used to explore the time-lapse data set and find examples of 4D anomalies. Between the multi-attribute sets a so-called match is