We studied the classification of seismic images and compared the performance of three groups of observers with different degrees of expertise in the seismic domain. The group of more experienced observers was more accurate when classifying a target fragment as belonging or not to a given seismic image. Observers of all groups classified high-relevance targets (i.e., those including geologically important features) more accurately than low-relevance targets (i.e., those without such features). However, the superiority of high-relevance targets did not increase as a function of expertise, as initially hypothesized. Rather, it was correlated with the individual skill in target classification.
