565
. «.^3
In: Wagner W., Székely, B. (eds.): ISPRS TC VII Symposium - 100 Years ISPRS, Vienna, Austria, July 5-7, 2010, IAPRS, Vol. XXXVIII, Part 7B
fruit plantations and study the relationships between remotely-
sensed indices and ground control measures of fruit trees. Based
on field work unique data of all pear trees, registered in the
genetic collection, were collected to one ESRI ARCGIS data
base. Height measurements and digital photos of the trees were
also carried out during the survey. Digital photos were taken of
the fruits as well, although earlier taken photos of genetically
typical, ripe fruits were set to the database. The pictures were
geotagged to air photo based on coordinates.
The n-dimensional hyperspectral image contained 359 channels.
The bad bands, having bad signal/noise rate were excluded from
further assessments (Tamas et. al 2005).
The characteristics of the reflectance curves of each pear tree
species result from the large amount of absorption of
chlorophyll content at 450 - 670nm wavelength intervals. On
the other hand, reaching the near infra-red (NIR) interval, the
reflectance of the healthy fruit tree leaves are raising markedly
at 700nm (Tamas, 2010). Besides, the reflectance value of the
vegetation without any stress is high at NIR intervals, but low at
red wavelength interval. The chlorophyll content is one of the
indicators of the state of health before ripening phenological
phase (Burai et al. 2009). The leaf samples taken with R-
Row/T-Tree positions in the study area were collected and
analyzed on the basis of the field measured spectral curves
(Table 2).
Pearl Species
R/T
NDVI
SRI
REP
chlorophyll
ftg/g
Napoca
19/
103
0.62
4.96
718.4
883.52
Márianosztrai
19/
113
0.41
4.95
720.7
509.12
República
19/
83
0.58
3.39
712.1
982.72
Ananász
(Ananas)
19/
79
0.69
5.26
710.9
4522.88
Nyári esperes
(summer deán)
19/
70
0.67
4.84
712.9
3960
Kiev
19/
109
0.489
2.79
721.7
407.36
Table 2. Example for chlorophyll content and spectral indices
The Pearson correlation between chlorophyll and Normalized
Vegetation Index (NDVI) was 0.8, between chlorophyll and
Simple Ratio Index (SRI) was 0.54, between chlorophyll and
Red Edge Position (REP) was -0.76.
During the survey NDVI index was calculated from the
hyperspectral image. The segmentation of the NDVI image was
based on the GIS database of field measurements. After
vectorising the obtained segments, those pear species having
different biomass weight were selected. Pixels covered by
canopies of these trees, could be spectrally clear and unmixed.
Although the results show, that real endmember pixels can only
be found at trees having several canopy level and at least 1 m
canopy diameter. Because of the replacement of necrotized trees
the spectral properties of grass zone between rows can provide
spectrally mixed values (as a 2 nd type errors of commission) due
to the possible underdevelopement of trees. On the other hand,
the spectral properties of neighboring trees can be mixed due to
the overdeveloped tree canopy having more than 4 m diameter,
which larger than the spacing in the row. Generally these errors
of ommission (1 st type) had minor role. The resulted, classified
spectral data could be applied as a spectral library, which are
suitable for the detailed examination of plant physiology and
spectral data. The optimal climatic conditions for pear are
generally cooler and humid, with at least 65% of relative air
humidity. There is a lack of these conditions in Ujfeherto at the
30-40% in a year. At the examined site, pear species with large
biomass concerned as those species which had major tolerance
against drier and unfavourable environmental conditions. These
genetic properties could be very important in the frame of the
potential climate change. Stress, caused by the relative water
lack, evolves in a short term physiological process. The first
symptoms of water stress are often not visually observed,
although those have negative effect on the yield quantity and
quality. The conventional invasive measurements can hardly be
reproduced, because of the sampling (cut) of the vegetation
tissue (i.g. leaf, shoot) or only few cm 2 of leaf area is measured,
which limits the representative sampling. The airborne
hyperspectral remote sensing data eliminate these
disadvantages, since it can provide detailed spectral data from
the whole canopy.
In Figure 2 the continuum removed spectra of Bajai, KorteB/3,
Szentendrei csaszar, Verteskozmai pear species are shown
descending order of biomass. Continuum Removal is to
normalize reflectance spectra to compare individual absorption
features from a common baseline. The continuum is a convex
hull fit over the top of a spectrum using straight-line segments
that connect local spectra maxima. The first and last spectral
data values are on the hull; therefore, the first and last bands in
the output continuum-removed data file are equal to 1.0. The
resulting image spectra are equal to 1.0 where the continuum
and the spectra match, and less than 1.0 where absorption
features occur.
Figure 2. The pear species have species specific spectral
properties
Based on the horticultural data, the four biomass segments
showed strong correlation with the water stress tolerance. The
pear tree species were grafted to the same wild pear rootstock,
thus there were no differences in the water consumption root
zone. In accordance with the spectral data, the “Bajai” species
produced large canopy with 5.7 m height and large yield
quantity (Figure 3).