Download details

Influência da Altura de Voo Sobre a detectabilidade do Grau de Infecção Ramularia areola em Culturas de algodão por Imagens Multiespectrais Adquiridas em Aeronave Remotamente Pilotada Influência da Altura de Voo Sobre a detectabilidade do Grau de Infecção Ramularia areola em Culturas de algodão por Imagens Multiespectrais Adquiridas em Aeronave Remotamente Pilotada

Orientador
Ano de Defesa 2017

ABSTRACT

The constant emergence of new technologies does not guarantee contributions to the

improvement of the man-nature relationship. There is a need for precise control over

the possibilities related to these technologies, so all the benefits involved can have

expected effects. The recent development of Remotely Piloted Aircraft (RPA), open up

growing perspectives of their adaptations in applications of Remote Sensing in

agriculture. Making it possible to investigate objects at different scales by analyzing

the ability to interpretate the same target in different spatial resolutions. The search is

an approach that involves several scales of Remote Sensing, basing on a multi-rotor

ARP, equipped with multispectral camera. Through statistical analysis (Boxplots and

Mann-Whitney Test), the sensor's ability to record the level of cotton involvement by

Ramularia areola and how the spectral readings behaves at different heights was

evaluated. The research area is located in the experimental field of the IMAmt (Instituto

Mato-grossense do algodão), located in the municipality of Primavera do Leste. The

municipality is one of the main cotton producers of the State of Mato Grosso. In the

2014/15 season, it was among the five largest cotton producers in the state. The

aircraft used in the survey was German-made Oktokopter model (MK ARF Okto XL

6S12) and the multispectral camera was the Tetracam ADC Lite. The experimental

field was divided into plots and were composed of two planting lines that have

individual levels of infection. In horizontal flight, images were taken at flight heights of

100 m, 300 m, 500 m and 700 m. Through a methodology developed by the IMAmt,

the infection levels were evaluated being reclassified in five levels to improve the

application of the statistics. The research demonstrates results in two categories, data

with and without normalization. The factors demonstrate the existence of external

effects interference with the increase of the image height. With the aid of statistical

software graphs and tables were created for analysis. The digitals numbers (ND and

NDcor) decreased in the bands of green and near infrared and increased in the red

band with increased level of infection characterizing them with behavior typical of

vegetation with infection. The increase in image height showed a tendency of

differentiation capacity in the levels of infection. However, the detectability of the cotton

infected by Ramularia areola (in data with and without normalization) presented,

statistically, a better separation in the distribution of the data between the levels of

infection, than referent to the image height.

Keywords: Cotton. Mapping. RPA. Scale. Ramularia areola

Information
Autor
Orientador Onélia Carmem Rossetto
Ano de Defesa 2017
Size 4.74 MB
Rating
(0 votes)
Created by Verônica
Changed by
Downloads 79
License
Price

Defesas

Não há eventos programados!