Search

Reliable estimation of the electro physical parameters of surfaces using radiometric equipment is an actual area of remote environmental monitoring. The disadvantage of the procedures developed today for assessing the indicated properties of surfaces according to their own thermal radiation, prevents their practical use.In the framework of the statistical theory of detection-measurement based on the results of processing radiometric signals during multichannel reception, an identification method is developed and a decision-making rule were synthesized in favor of any of the hypotheses put forward about the values of the parameters of the observed surface.The methodology for solving the problem and the inherited identification rule meet the criterion of maximum likelihood. A distinctive feature of the synthesized decision-making rule: critical statistics based on the eigenvalues of the covariance observation matrix under multichannel reception conditions. The implementation uses standard computing operations.Moreover, a distribution density of critical statistics obtained because tabulated solving the problem. This circumstance makes it possible to set the required value of the significance level “apriori”, therefore, the value of the comparison threshold.Certification of the synthesized test is the rules for deciding in favor of the hypothesis about a given value of the components of the Fresnel reflection coefficient. It was carried out at the level of digital statistical modeling. The significance level was set equal to 0.05 and 0.1. Based on the simulation results, there were analyzed the dependence of the number of correct decisions (the fulfillment of the corresponding hypothesis) on the signal-to-noise power ratio for a given angular parameter of the signal.The performance characteristics obtained at the level of digital statistical modeling of the developed test confirmed, firstly, the conclusions of theoretical studies. Secondly, they give specialists in the field of development of radiometric systems the opportunity to evaluate both the effectiveness of synthesized tests at various numerical values of the probability of an error of the first kind and the complexity of their practical implementation.