The article discusses a new method for estimating parameters in a 2-D sinusoidal model, which is used in signal processing and texture analysis. The traditional least squares estimators work well with noise but struggle with outliers. The proposed weighted least squares estimators are more robust in the presence of outliers, behaving similarly to other robust estimators like least absolute deviation estimators. The researchers found that the weighted least squares estimators have consistent and normal properties, and simulations show that they are effective in practice.