A generalization of orthogonal regression estimators is considered for estimation of simple regression coefficients in the error-in-variables model with observations from a mixture with varying concentrations. Consistency and asymptotic normality of theestimators are shown. The dispersion matrix is evaluated.
