LAUSR

It is of vital importance to search for an objective quality metric which can provide insights into the geometric strength of the underlying adjustment model. The Minimal Detectable Bias (MDB) measure has been extensively used in geodesy and photogrammetry. However, MDB is a hypothetical quality measure. To this end, the Maximum Tolerable Bias (MTB) metric is introduced in this contribution. The definition of MTB measure is independent of the distribution of observables, the test statistics, the specified significance level and the given detection power for the one-sided hypothesis test. Consequently, MTB is an objective quality measure for least-squares adjustment models. Examination of the scaled version of MTB can reveal the hardly detectable data points. All the hardly detectable points should be removed before performing conventional outlier detection procedures.