Random sample consensus
Random sample consensus (RANSAC) is an iterative method to estimate parameters of a mathematical model from a set of observed data that contains outliers, when outliers do not affect the values of the estimates. Therefore, it also can be interpreted as an outlier detection method.