Сравнение алгоритмов дешифрации аэрофотоснимков земель сельскохозяйственного назначения

Abstract

A comparison of two algorithms, calculation of Euclidian distance and Mahalanobis distance, is presented. The purpose of this comparison is to increase the accuracy of decoding of aerial photographs used in evaluation of agricultural area and green vegetation conditions. It is shown that it is necessary to take into account the spectrum influence of neighboring points and also to include into the class more than one learning sample, using several reference fields in order to increase the classification accuracy of images, obtained by super-light flying objects at the height of 1–3 km, and which has three channel of visible spectrum without use of the infrared channel.