Algorithm of objects classification by optoelectronic systems of unmanned aerial vehicles

In this work we present the algorithm of object classification on the basis of thermal images obtained by optoelectronic systems of unmanned aerial vehicles. In particular, we show that the searching of objects by the distribution of estimated values of the thermal conductivity is more effective since it allows to divide the objects into classes due to their difference in thermal inertia. Numerical estimates of detection probability for each class and the total error of detection are presented.