Summary

The aim of this thesis was pair the two data source used in precision farming: remote sensing and sensor data from the WSN. To do this it was necessary to be familiar with these data sources and their use in precision agriculture. Both with their pros and cons.

Remote sensing offers the opportunity to obtain complete data from the entire field at one time. Different platforms, from satellite to the UAV, also allow to obtain data almost in a user predefined parameters. Use of multispectral and hyperspectral cameras allow examine especially the variability of vegetation. For the monitoring of the vegetation remote sensing methods have many advantages from obtaining information from the entire field at one moment to the opportunity to make such measurements more frequently and cheaper than conventional methods allow. Variability of soil monitoring using remote sensing methods has limited use, because it sensing only the surface and the strong influence of soil moisture and limited time range appropriate for imaging (surface must be covered with vegetation).

For monitoring of soil parameters, such as soil moisture, nutrients and soil pH, today more use various types of sensors from stably positioned, monitoring mainly soil moisture, to the sensors located at different moving pltaform. These sensors use different physical properties of soil.

WSN could be considered as special type of permanently mounted sensors. They besides monitoring soil properties can monitor many other phenomena. Especially in the arboriculture WSN is mainly used for monitoring of meteorological phenomena, for example, for vegetation protection against frost phenomena, which could adversely affect the crop.

In this work we were used multispectral image data from the camera Tetracam ADC borne UAV system Hexacopter. To the thus obtained images was designed processing procedure which included mainly the image correction methods to achieve more relevant DN values. On the basis of this procedure has been prepared for the program ArcGIS toolbox with tools for the preparation of correction data and the subsequent batch process. From sensor data was used wireless sensor network based on the elements offered by Libelium. These were mainly meteorological sensors (temperature, humidity, rainfall, wind direction and speed) and sensors for monitoring the composition of the atmosphere (CO2, NO2 and other gases). Data obtained from this network had to be adjusted to be in a form suitable for storage. It is necessary to remove or repair the faulty records and parsing of data on individual values. To the resulting data were prepared scripts which allow batch import of both types of data to the database.

It inherently produces these data is a very different dates. Whose pairing in one data source is difficult. The database must contain both sensor data, which are point character and image data. Including descriptive information about sensors and network nodes (position and characteristics of individual sensors), as well as descriptive information about the images and characteristics of cameras used.