Recognition of Natural Objects on Air Photographs Using Neural Networks
M. N. Favorskaya, N. Y. Petukhov
Reshetnev Siberian State Aerospace University
favorskaya@sibsau.ru, n_petukhov@sibsau.ru
Keywords: recognition of landscape images, fractal parameters, artificial neural networks
Pages: 34-40
Abstract
This paper presents a method for recognizing natural objects on air photographs based on a two-level segmentation procedure and a complex calculation of statistical and fractal parameters using artificial neural networks to determine the category of a natural object and then the type of the object in the category determined. This method is effective for recognizing tree species on air photographs. The basic formulas for calculating the characteristics of natural textures are given, and the topology of the artificial neural networks used is substantiated.
|
