Long-range scanning probe microscope (SPM) measurements are usually extremely time consuming as many data need to be collected, and the microscope probe speed is limited. In this article, we present an adaptive measurement method for a large-area SPM. In contrast to the typically used line by line scanning with constant pixel spacing, we use an algorithm based on several levels of local refinement in order to minimize the amount of information that would be useless in the data processing phase. The data obtained from the measurement are in general formed by xyz data sets that are triangulated back with a desired local resolution. This enables storing more relevant information from a single measurement as the data are interpolated and regularized in the data processing phase instead of during the measurement. In this article, we also discuss the influence of thermal drifts on the measured data and compare the presented algorithm to the standard matrix-based measuring approach.