Аннотации:
This article is about methods of analyzing aerial images. Images from
Planet.com for crops in North Kazakhstan owned by the Center for Cereal
Production and Research. A.I. Barayev. The main goal of the research work
is to develop and implement algorithms that allow identifying and
distinguishing factors in aerial photographs that adversely affect the growth
of plants during the growing season. Spectral brightness coefficient (SBC),
normalized difference vegetation index (NDVI), textural features, clustering,
and integral transformations are used to solve the problem. Particular attention
has been paid to the development of software tools for selecting features that
describe textural differences to divide texture regions into subregions. That is
weeds, and pests in aerial images. The application of a set of textural features
and orthogonal transformations to the analysis of experimental data is
explored to identify regions of potentially correlated features in the future.
The analysis of the received data made it possible to determine the
characteristics of changes in the reflective capacity of agricultural plants and
weeds in certain stages of the growing season. The obtained information is of
great importance for confirming the observations from space remote from the
aerial images.
