Detekcja posuszu kornikowego z wykorzystaniem zobrazowań BlackBridge na przykładzie drzewostanów Sudetów i Beskidów.
PBN-AR
Instytucja
Instytut Badawczy Leśnictwa
Informacje podstawowe
Główny język publikacji
pl
Czasopismo
Sylwan
ISSN
0039-7660
EISSN
Wydawca
POLSKIE TOWARZYSTWO LESNE
DOI
Rok publikacji
2016
Numer zeszytu
9
Strony od-do
707-719
Numer tomu
160
Link do pełnego tekstu
Identyfikator DOI
Liczba arkuszy
Autorzy
(liczba autorów: 3)
Pozostali autorzy
+ 1
Autorzy przekładu
(liczba autorów przekładu: 0)
Słowa kluczowe
pl
BlackBridge
condition of forests
satellite image classification
mountain areas
Streszczenia
Język
en
Treść
BlackBridge imagery is one of the new means of information used in forest condition analysis. Rapid Eye satellite data with a 5 m spatial resolution register spectral information from 440 to 850 nm through 5 spectral bands. This range of electromagnetic spectrum provides information on plant chlorophyll content as well as cell structure. Such data allows to monitor vegetation condition. This paper focuses on a research conducted in the Sudety and Western Beskidy mountains (southern Poland). The aim of the research was to verify whether high resolution satellite imagery is applicable in detection of the damages caused by Ips typographus and acid rain in Norway spruce dominated stands through supervised classification. BlackBridge Rapid Eye satellite images from 2012 and 2013 were analysed. Various modifications of classification methods were tested, including change in combination of spectral bands. Each method resulted in different classification accuracy. Best results were observed in case of the Maximum Likelihood classification method applied on all spectral bands. The analysis showed that the time of the image registration has a significant impact on classification results. The average classification accuracy for 2012 images was 0.53, whereas for 2013 – 0.69. Moreover, information gathered from 5 m pixels is too general to classify individual dead trees in a precise manner. Tested methods are applicable only in detection of clusters of dead trees.
Cechy publikacji
ORIGINAL_ARTICLE
Inne
System-identifier
755921