Geometry accuracy of DSM in water body margin obtained from an RGB camera with NIR band and a multispectral sensor embedded in UAV

Abstract and Figures

The photogrammetry techniques are known to be accessible due to its low cost, while the geometric accuracy is a key point to ensure that models obtained from photogrammetry are a feasible solution. This work evaluated the discrepancies in 3D (DSM) and 2D (orthomosaic) models obtained from photogrammetry using control points (GCPs) near a reflective/refractive area (water body), where the objective was to evaluate these points, analysing the independence, normality and randomness and other basic statistic. The images were obtained with a 16 MP Canon PowerShot ELPH 110S with a modified NiR band and a multispectral sensor Parrot Sequoia, both embedded in a hex-rotor UAV in flight over the Unisinos University’s artificial lake in the city of São Leopoldo, Rio Grande do Sul, Brazil. Due the distribution of the data found to be not normal, we applied non-parametric tests Chebyshev’s Theorem and the Mann–Whitney’s U test, where it showed that the values obtained from Sequoia DSM presented significant similarities with the values obtained from the GCP’s considering the confidence level of 95%; however, this was not confirmed for the model generated from a Canon camera, showing that we found better results using the multispectral Parrot Sequoia.
Content may be subject to copyright.
Full Terms & Conditions of access and use can be found at
European Journal of Remote Sensing
ISSN: (Print) 2279-7254 (Online) Journal homepage:
Geometry accuracy of DSM in water body margin
obtained from an RGB camera with NIR band and
a multispectral sensor embedded in UAV
Dalva Maria de Castro Vitti, Ademir Marques Junior, Taina Thomassin
Guimarães, Emilie Caroline Koste, Leonardo Campos Inocencio, Maurício
Roberto Veronez & Frederico Fábio Mauad
To cite this article: Dalva Maria de Castro Vitti, Ademir Marques Junior, Taina Thomassin
Guimarães, Emilie Caroline Koste, Leonardo Campos Inocencio, Maurício Roberto Veronez &
Frederico Fábio Mauad (2018): Geometry accuracy of DSM in water body margin obtained from
an RGB camera with NIR band and a multispectral sensor embedded in UAV, European Journal of
Remote Sensing, DOI: 10.1080/22797254.2018.1547989
To link to this article:
© 2018 The Author(s). Published by Informa
UK Limited, trading as Taylor & Francis
Published online: 26 Nov 2018.
Submit your article to this journal
View Crossmark data
Geometry accuracy of DSM in water body margin obtained from an RGB
camera with NIR band and a multispectral sensor embedded in UAV
Dalva Maria de Castro Vitti
, Ademir Marques Junior
, Taina Thomassin Guimarães
Emilie Caroline Koste
, Leonardo Campos Inocencio
, Maurício Roberto Veronez
and Frederico Fábio Mauad
Faculty of Technology of Jahu, Environment and Water Resources, Jau-SP, Brazil;
EESC, University of Sao Paulo, Sao Carlos-SP, Brazil;
Polytechnic School, Unisinos University, São Leopoldo-RS. Brazil
The photogrammetry techniques are known to be accessible due to its low cost, while
the geometric accuracy is a key point to ensure that models obtained from photogram-
metry are a feasible solution. This work evaluated the discrepancies in 3D (DSM) and 2D
(orthomosaic) models obtained from photogrammetry using control points (GCPs) near
a reflective/refractive area (water body), where the objective was to evaluate these
points, analysing the independence, normality and randomness and other basic statistic.
NiR band and a multispectral sensor Parrot Sequoia, both embedded in a hex-rotor UAV
in flight over the Unisinos Universitys artificial lake in the city of São Leopoldo, Rio
Grande do Sul, Brazil. Due the distribution of the data found to be not normal, we
applied non-parametric tests Chebyshevs Theorem and the MannWhitneysUtest,
where it showed that the values obtained from Sequoia DSM presented significant
similarities with the values obtained from the GCPs considering the confidence level
of 95%; however, this was not confirmed for the model generated from a Canon camera,
showing that we found better results using the multispectral Parrot Sequoia.
Received 30 December 2017
Revised 5 November 2018
Accepted 10 November
Accuracy; SfM; UAV;
mapping; geo-statistics;
DSM; photogrammetry
The 3D modelling generated from images acquired
from high-resolution cameras embedded in
unmanned aerial vehicles (UAV) is consolidating as
an alternative technique with low cost in large-scale
mappings (Smith, Carrivick, & Quincey, 2016).
The Structure from Motion (SfM) technique with
the Multi-View Stereo algorithm aims to reconstruct
a surface or object from the matching of common
points collected from several images, where each
point consists of a position and a colour extracted
from an image (Debevec, Taylor, & MaliK, 1990;
Fonstad, Dietrich, Couville, Jensen, & Carbonneau,
2013; Oliveira, 2002; Remondino, Scaioni, & Sarazzi,
2011; Snavely, Seitz, & Szeliski, 2008). This method
does not require metric cameras, giving the SfM
a status of a more feasible approach, due to the
cheaper price and availability of this type of the non-
metric cameras (Smith et al., 2016).
For an efficient construction of the 3D surface is
important that the angular separation between the
images (considering the position of the camera dur-
ing the capture) does not exceed 2530 degrees,
achieved with a minimum of 6080% of images over-
lapping in an individual location. Also requiring good
photo quality for a better performance of the detect-
ing algorithms (Smith et al., 2016; Verhoeven, 2011).
The SfM algorithm generates a sparse point cloud
using a local reference system. To georeferencing the
point cloud we need distributed points referenced
with a global navigation satellite system (GNSS) to
generate correct metrics (scale and global position)
(Lobnig, Tscharf, & Mayer, 2015). We need at least
four control points in a global system to obtain an
equivalent metric to the real-world scene. The model
in an absolute reference system is achieved with the
7-parameter Helmerts similarity transformation
(Lerma, 2002; Lobnig et al., 2015).
The next step is to create a dense point cloud that
is generated with the Multi-View Stereo (MVS) algo-
rithm that calculates the geometry pixel-by-pixel,
reproducing higher level of detail in the scene
(Furukawa & Curless, 2010; Kraus, 2004).
After the dense point cloud creation, we are able to
build the proper 3D model applying triangulation
algorithms to build a mesh, also with the camera
positions and interior orientation parameters we
have the orthophotos, where all objects with
a certain height are accurately positioned in a 2D
plane (Verhoeven, 2011).
CONTACT Ademir Marques Junior Polytechnic School, Unisinos University, São Leopoldo-RS. Brazil
© 2018 The Author(s). Published by Informa UK Limited, trading as Taylor & Francis Group.
This is an Open Access article distributed under the terms of the Creative Commons Attribution License (, which permits
unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.
The first generated 3D model is the digital elevation
model (DEM) or the digital surface model (DSM) that
considers the elevation of all elements above the
ground. The proper georeference of this model give
us the correct global position and scale of the model.
The digital terrain model (DTM) is a 3D model
that does not consider elements above the ground,
e.g. trees and buildings. To achieve this, filters are
applied to remove these elements, as this process can
be repeated by varying the reclassification parameters
to obtain a cleaner DTM. As final products of SfM,
we have the DSM or DTM (3D models), and the
orthomosaic (2D image).
To evaluate the positional accuracy of products from
SfM + MVS most works have in common the compar-
ison with differential GNSS control points, while others
compare with laser scanner data. In addition, the root
mean square error is the most used statistical value.
Another important aspect of the statistical evaluation is
the predominance of studies only considering parametric
statistics (normal distribution) and the analysis centred
in the horizontal plane (2D). (Harwin & Lucieer, 2012;
Joaquim Höhle, 2009;Laliberte,Herrick,Rango,&
Winters, 2010;Netoetal.,2017; Turner, Lucieer, &
Christopher Watson, 2012).
Each country regulates the mapping process differ-
entially. Most of them recommend at least 20 control
points and assume that the sample has a normal
distribution (Atkinson, 2005; Ariza & Atkinson,
2008). In addition, the standards do not detail the
steps of assessment of accuracy and whether the
assessment of horizontal (positional) accuracy is per-
formed independently Xand Y(linear error) or joint
XY (circular error). The Zcomponent (height) is
always evaluated independently (Ariza & Atkinson,
2008). Our work uses Efor X,Nfor Yand h
for Zin UTM coordinates in meters (m).
An additional consideration is, that we did not
find works that consider large reflective areas as
lakes and others types of water bodies, this research
aimed to evaluate the positional accuracy of products
generated by the SfM and MVS for mapping pur-
poses using an RGB (with modified near infrared
band) and a multispectral sensor.
This work is presented as follows: the methods and
material, workflow and the statistical basis; the study
over the artificial lake at the Unisinos University,
where the products of the SfM + MVS process were
evaluated; the discussion of results of this work and
how they compare with similar works; and lastly the
conclusion of this work were we establish some
remarks and considerations for this and future works.
Materials and methods
This section describes the methods and materials
used in this work. First with the equipment used to
acquire the images and second describing the statis-
tical techniques to evaluate the products obtained
from the sensors.
Studied area
Following the premise of evaluating the model accu-
racy around a water body, the area chosen for this
work was the artificial lake presented in the campus
of the Unisinos University in the south of Brazil in
the state of Rio Grande do Sulin the city of São
Leopoldo. The selected area of evaluation is marked
in Figure 1.
The studied area was chosen in order to evaluate
the influence of a reflective/refractive area near the
points used for assessment localized in the water
body margin and the use of two different types of
cameras/sensors and consequently to indicate the
better camera in this situation.
This work followed the steps presented in Figure 2,as
some of the processes are detailed in the next
Image acquisition
83 images were collected for both the Canon
PowerShot camera ELPH 110HS 16.1 MP (Toyo,
Japan) (15 cm GSD and focal distance of 4.4 mm)
with modified NIR band and the multispectral sensor
Parrot Sequoia 16 MP (Paris, France) (14.52 cm
(11 cm) GSD and focal distance of 4.88 mm) in flight
over the studied area with height of 120 m above the
ground. Both camera/sensors were shipped in the
hex-rotor UAV ST800, as shown in Figure 3.
Structure from motion processing
In this phase, the images collected are combined with the
flight log into specific software that uses computational
vision and the algorithm Structure from Motion (SfM) to
match common points in the images using points of same
colour to recreate the scenes in a 3D model (the digital
terrain model) or in a flat surface model (orthomosaic).
The software Pix4Dmapper (Lausanne, Switzerland) was
used to perform this task.
Statistical analysis
The verification of the positional accuracy should be
performed independently for the horizontal compo-
nents (Eand N) and the vertical component (h),
where Eand Nare the positions in UTM coordinates,
and his the soil elevation (Atkinson, 2005; Ariza &
Atkinson, 2008).
To assess the accuracy, the coordinates of the
points identified in the orthomosaic (EN) and in the
3D model (h) need to be compared with their homo-
logues extracted from a source at least three times
more accurately (Preciado, 2000), (the most accurate
technique is the GNSS differential survey), and the
minimum number of checkpoints recommended is
20 (ASPRS, 2015; Brazil, 1984; Federal Geographic
Data Committee, 1998) to 167 as in STANAG:2215.
The first step is the calculation of the discrepancies
(horizontal and vertical) e
between the coordinates
from the orthomosaic (2D positional or EN) and
Figure 1. Unisinos University satellite view.
Figure 2. Flowchart of research activities for positional accuracy analysis of photogrammetry from UAVs and SfM processing.
Figure 3. Aircraft ST800 with the cameras Canon ELPH 110 HS and the Sequoia sensor.
the 3D model (combining the three axes E, N and h),
against the coordinates of control C
obtained by
a more precise method (Atkinson, 2005; ASPRS,
2015; Federal Geographic Data Committee, 1998;
Preciado, 2000) as in Equation (1):
This work uses the positions as EN coordinates, as e
values obtained are the distances in meters between
the control coordinates obtained with differential
GNSS and the control points identified in the geor-
eferenced orthomosaic. The discrepancies in height
(h) or soil elevation are measured using the elevation
obtained with the differential GNSS in the control
points against the elevation found in the control
points identified in the georeferenced 3D model.
Following, the inspection of the atypical points or
outliers is performed (Preciado, 2000). These points
are not necessarily coarse errors but must be investi-
gated. We need to estimate the expected error (σ
basing on the steps of data acquisition and image
Estimating the expected error
The expected error value must consider all possible
errors during the phases of data acquisition and pro-
cessing (Brazil, 1984). In 3D Modelling with UAV the
expected error, σ
, for the image-processing phase
can be estimated by Equation (2), where its elements
are described in Table 1.
Then, the σ
error is multiplied by 3 and the
result obtained is compared with the absolute discre-
pancies at each point, where values that exceeded this
value are excluded from the analysis. The points
excluded should be between 10% and less than 20%
of the sample (Bustos, 1981). After the inspection for
outliers, the study of previous hypotheses of indepen-
dence, normality and randomness is carried out.
Independence test of sample discrepancies
This test is done after the exclusion of outliers and
before the normality test, to verify the correlation
between the discrepancies and the distances between
the checkpoints, and properly select the most ade-
quate statistical tests for each case.
The independence test of the data can be made by
geostatistical analysis (Santos, 2015; Santos, Medeiros,
Santos, & Lisboa Filho, 2017), using the experimental
semivariogram. The experimental semivariogram for
each positional discrepancy (EN and h) are obtained
from the calculation of the semivariances σGNSS UAV
given by Equation (3), where N(h) is the number of
positional discrepancy values pairs of dp(x
) and
+h) separated by a distance of h.
dp xi
ðÞdp xiþhðÞ½
The experimental semivariogram is represented by
a chart, where the ordinate is the semivariance, and
the abscissa is the distance between the elements of the
sample. The correlation occurs until the point where
the semivariance of the discrepancies stabilizes in the
semivariogram, making the RANGEdistance, and
after that point, the discrepancies are independent
(Santos, 2015; Yamamoto & Landim, 2013). In Figure
4, in (a) we observe an adjusted theoretical model, with
all elements of a semivariogram, RANGE,NUGGET
EFFECTand SILL; In (b) there was no correlation
between the discrepancies (only the NUGGET
EFFECTis present) indicating independence.
If the independence is verified, the analysis of
normality and randomness, followed by the analysis
of accuracy is performed. Otherwise, the theoretical
model can be adjusted to exponential, spherical or
Gaussian for semivariograms with SILL(Santos,
2015; Soares, 2000; Yamamoto & Landim, 2013).
After the semivariogram modelling, the adjust-
ment is done by techniques such as ordinary least
squares, weighted least squares, maximum likelihood
and or restricted maximum likelihood. The data are
cross-validated generating the standardized residue
(independent, normal, non-tendentious and homoge-
neous), and the correlation coefficient R
is deter-
mined. Then, the correlation of the standardized
residuals with the positional discrepancies is verified.
If the correlation is greater than 0.6 (strong correla-
tion), it is followed by the verification of the accuracy
based on the standard residue, if it is less than 0.6
(weak correlation), the semivariogram should be
revised (Santos, 2015; Santos et al., 2017).
To verify the independence of data using the semi-
variogram, we need to analyse the pairs of points in
multiple directions (directional semivariogram). This
is done to verify the independence between the loca-
tion of the points to the distance and the geographical
direction (in relation to north). Considering this, this
work used the directions 0°, 45°, 45°, 60° and 90° to
analyse the correlation in multiple directions, and
a length of 30 m for lag increment, that is, the
distance interval to be considered as we approach
the cut (maximum) distance in each direction
σpriori ¼σ2
sensor þσ2
support þσ2
image support þσ2
processing þσ2
DEM þσ2
(1000 m in this case), and an Rvalue of 10 (tabulated
value) for both EN and hdiscrepancies.
Normality tests
The null hypothesis in the normality test is that the
discrepancies are distributed in a Gaussian function
(ASPRS, 2015; Brazil, 1984; Federal Geographic Data
Committee, 1998); however, most accuracy standards
for cartographic products do not consider free dis-
tribution. ASPRS (2015) standards recommend per-
forming accuracy tests for data that are not in
a normal distribution.
The EN position data usually have a normal dis-
tribution from 20 sample points, so parametric tests
and the estimators, mean, standard deviation and
RMSE satisfy them. As this distribution occurs in
less frequency for the elevation discrepancies, which
usually has a free distribution, so the previous
hypothesis must be based on non-parametric meth-
ods based on the median (Atkinson, 2005).
The ShapiroWilk test was selected for the nor-
mality test for having a better performance compared
with KolmogorovSmirnov, Chi-
square and Studentst-test, when evaluating free dis-
tribution data (Neto et al., 2017; Santos, 2015;
Torman, Coster, & Riboldil, 2012).
The previously established hypothesis considering
a significance of 0.05 is described below:
: the distribution of the discrepancies EN or
hfulfil the normal function.
: the distribution of the discrepancies EN or
hdo not fulfil the normal function.
If the previous hypothesis (H
) is confirmed, the
mean and the standard deviation of the sample can be
attributed to the population and the horizontal and
altimetric accuracy can be verified by the mean, the
standard deviation or the RMSE. On the other hand, if
the data are independent and the distribution function
is unknown, or if it is known as non-normal, the central
limit theorem (CLT) can be applied (Santos, 2015).
Otherwise, if we have free distribution, the
Chebyshevs theorem could be applicable as it can
accept any shape of distribution and is valid for
akfactor (multiplier of standard deviation) greater
than 1, where at least (11/k
) of the data values are
positioned within of the limits of ktimes the stan-
dard deviation in relation to the arithmetic mean
(Mann, 2008). This theorem allows estimating if the
points are within a range for a free distribution. For
example, for a threshold of twice (kvalue) the stan-
dard deviation is assumed that at least 75% of the
data positioned within this range, while in the normal
distribution, it is assumed that 95% of the data would
be in that range.
Randomness tests
The most used test to verify the randomness is the
Runs Test, representing a series with one or more
consecutive occurrences of the same result in which,
there are only two results (Atkinson, 2005; Mann,
2008; Torman et al., 2012). For this analysis, the
previous hypothesis is declared as follows:
the discrepancies are randomly distributed.
the discrepancies are not randomly
To perform this test, the median is used as
a parameter as below:
discrepancy median.
discrepancy < median.
The discrepancies are separated into two groups:
to the positives to H
and n
to the answers that
follow H
. After that, the number of arrangements
designated by Ris counted as each event of a chain of
yesor no(e.g. YYNNNYNYY, R= 5). Therefore,
Rrepresents a statistic with its own sample distribu-
tion, where its critical values are determined in two
Table 1. Error elements considered in the calculated global expected error.
Error element Description
Mean positioning error of the images in function of the embedded GNSS receptor technology, which will be
considered only for the models directly georeferenced.
Error in the camera calibration, given by the differences between the initial coordinates of the centre of the image
and the coordinates optimized by the camera.
Positioning error of the control points collected in the field; extracted from the post-processing report.
σ2D support ¼σ2
0:5For the analysis of the horizontal accuracy.
σ3D support ¼σ2
0:5For the analysis accuracy of the DTM.
Identification error of key points, usually caused by flight characteristics, weak photo overlap and the software
performance. It can be reduced by selecting better quality images or using filters, in this case, one must
consider the efficiency of filtering anomalous or undesirable points.
Identification error of the points of support in the image, function of the resolution of the targets. This study
adopted a value of 3 times the image spatial resolution.
Interpolation error due to filter failure in discretization of dense cloud layers. Has a stronger influence on the
Zcomponent, since not separating points from the terrain of other features, such as trees and buildings, the
filtered model of the DSM could be a little different from the DTM.
Error of identification of the control points in the cartographic product, function of the final spatial resolution of
the format of the check elements, whether artificial targets or existing structures. Similarly, to the σ2
image support,
in this study was adopted the resolution of 3 times the spatial resolution of the orthomosaic.
For values of n
and n
lesser than 15, we use the
table of critical values for a two-tailed test equal to
0.05. This table presents two critical values (c
and c
in function of n
and n
. These values represent the
limit values to accept the initial hypothesis H
For values of n
and n
greater than 15, we can use
the normal approximation, where the Zvalue is cal-
culated in function of the mean of R, μRand from the
standard deviation of RσR, given in Equations (4),
(5) and (6).
μR¼ð2:n1:n2=n1þn2Þþ1 (5)
With the Zvalue and establishing a confidence level
(0.01 < a< 0.05), the standardized normal distribu-
tion table is searched for the critical Zvalue, and then
a decision is taken accepting or not accepting the null
The median test shows as a result, that at least
50% of the points tested fit into the null hypothesis
with a confidence level of 9599%, according to the
established value, attesting the sample randomness.
Descriptive statistical measures
After the tests of the previous hypotheses of indepen-
dence, normality and randomness of the sample dis-
crepancies of EN and h, the standard deviation,
(Equation 7) and RMSE
(Equation 8) were
where e
= discrepancy in the position EN or in the
height h;
ec= mean discrepancy in the position EN or
in the height h; n = number of tested points.
Other useful statistics are the mean referring to the
average value of a sum of numbers divided by the
quantity of values (for the arithmetic mean), the
median indicating the most central value in the sam-
ple (not easily distorted by outliers), and the CV or
the Coefficient of variation that is the standard devia-
tion divided by the mean resulting in a percentage
indicating the sample variability in relation to the
MannWhitneysU test
The MannWhitneysUtest is a non-parametric test
correspondent to the StudentsT-test applicable to
verify how similar are two sample groups based on
the median. The null hypothesis premise is that the
median of the populations is equal, and the alterna-
tive hypothesis is that the medians are different.
In this test, the values from the two samples are
joined and sorted in crescent order alongside the
ranked position for each value, then the ranked posi-
tion values are summed for each sample, then in the
equation below we calculate the Uvalue for each
sample using Equations (9) and (10):
where U
and U
are the calculated Uvalues; n
are the number of entries for each sample; R
are the sum of the ranked values for each sample.
The lesser Uvalue found is compared against the
Figure 4. Elements of a semivariogram for data with spatial dependence (a) and for spatial independent data (b).
table of critical Uvalues given according to the con-
fidence level. In this work we used the values found
in the models generated (E, N and hvalues) com-
pared with its pair obtained with the ground control
points (GCPs) values obtained with the differential
RTK-GNSS, using the table for Ucritical values for
a confidence level of 95%.
If the value found in the table of Ucritical values is
smaller than the calculated Uvalue the null hypoth-
esis is accepted and we can affirm that the two sam-
ples are similar (Mann, 2008).
Study case: Unisinos Universitys artificial lake
Following the workflow presented in Figure 2,83
images were collected in both sensors in a unique
flight. Alongside the image acquisition, we performed
the collection of 31 GCPs to be used in the statistical
evaluation, where six of this points were also used for
georeferencing the products obtained from the
SfM + MVS method. Figure 5 shows the points
extracted as control points used for georeferencing,
that were collected using a GNSS RTK marked in the
vertical signalling presented in the perimeter of the
SfM image processing
The image processing was performed in the
Laboratory of Advanced Visualization and
Geoinformatics in the Unisinos University. They
were imported to the software Pix4Dmapper Pro
version 3.1.23 alongside the flight log, position and
RGB of the central pixel in the images and the height
of the flight (yaw, pitch, roll).
The products obtained in the processing stage are
the 2D models (orthomosaic) for the Canon camera
(Figure 6) and the Sequoia multispectral sensor
(Figure 7), and the 3D models obtained from the
Canon camera (Figure 8) and the Sequoia sensor
(Figure 9) image processing.
Figure 5. Control points used for georeferencing and for the statistical analysis.
Figure 6. Orthomosaic obtained from the Canon camera.
Figure 7. Orthomosaic from the Sequoia sensor.
Figure 8. DSM obtained from the Cannon camera.
Figure 9. DSM obtained from images from the Sequoia sensor.
Positional accuracy verification
The positional accuracy was performed for EN (cir-
cular position) and hfor the 31 homologous points
identified in the orthomosaic and collected in the
field with GNSS Differential RTK. The Eand
Ncoordinates were extracted from the orthomosaic
using the sample point tool in QGIS 2.8.3 Wien.
Likewise, the altitude hwas read from the DEM/
DSM. Then a table with the position of all points
was exported with the CSVextension and tabulated
along with their counterparts.
Estimating the σ
The perceived error in each stage of the data acquisi-
tion, calculated by the Equation (1), in terms of the
resultant GSD (~6 cm) and the propagation of the
a priori error during the processing of the carto-
graphic products are shown in Table 2.
For the outlier inspection, the value of 2.5 times
the σ
was adopted instead of three times, with the
objective to reduce the error range in the final sam-
ple, but the same quantity of anomalous points was
excluded, resulting in three anomalous points both in
EN (circular error) and in h(altimetry), falling in the
expected rate of outliers according to Santos (2015).
We suspect that these outliers were generated due to
the fact that these GCPs were positioned near a tall
building, creating a more shadowed area with lower
visibility compared to the rest of the scanned area.
Analysing the positional data in Figure 10 for both
Sequoia (a) and Canon (b), the worse result are
clearly visible for the points 25, 26 and 27 (our out-
liers) for the data for both sensors, while the points 7
and 28 are almost in the error limit for the Canon 2D
With the discrepancies in EN and hinspected, the
analysis of previous hypotheses of independence,
normality and randomness of the sample is done.
Verifying the previous hypothesis
The independence of the EN and hdiscrepancies was
verified by geostatistical analysis, as reported in the
previous section. The point spacing and the semivar-
iograms of the EN discrepancies are shown in Figure
11 for the data obtained from the Canon camera,
where we observed no correlation between the dis-
crepancies and the distances (no dependence), with
the discrepancies for hpresenting a similar result also
for the Sequoia sensor.
The normality of the EN and hdiscrepancies for the
Canon and Sequoia sensor samples was verified by
analysing the discrepancy histogram compared to the
Figure 10. Positional data of the GCPs and the points extracted in the orthomosaic from Sequoia (a) and Canon (b).
Table 2. The resultants GSD values for the errors considered.
Sensor σ
(m) σ
(m) σ
(m) σ
(m) σ
support image
(m) σ
(m) σ
(m) σ
orto/DTM (m)
CANON 0.000 0.000 0.049 0.553 0.235 1.050 0.450 0.73/1.28
Sequoia 0.000 0.028 0.030 0.230 0.235 1.050 0.450 0.56/0.42
10 D. M. D. C. VITTI ET AL.
ShapiroWilk probability distribution function. The
p-value of the normality test by the Shapiro-Wilk func-
tion was below the threshold of 0.05 confirming that
the null hypothesis was rejected in all cases; hence, the
data used do not have a normal distribution.
The randomness test of the EN discrepancies was
performed applying the runs test considering the
median as base to the hypothesis of randomness.
This hypothesis was accepted for all samples.
Facing the results of the previous analysis we con-
cluded that the discrepancies for EN and hfor both
sensors (Canon and Sequoia) were independent and
random but with the distribution not normal indicat-
ing that we must proceed with non-parametric tests.
Descriptive statistics measures comparison and
inference tests
The descriptive statistic measures observed for the
orthomosaic and the DSM for both Canon and
Sequoia sensors are presented in Table 4.
We analysed three variables considering; the cir-
cular error or the 2D positional accuracy for
orthomosaic, which indicates torsion in the model;
the error in the altimetry for the 3D model; and the
3D positional error in the 3D model regarding the
three axes. This table shows better results in general
(lower values) for the Sequoia sensor in comparison
against the modified Canon cameraT (Table 3).
As the analysed data does not have a normal distribu-
tion, the mean and the standard deviation cannot repre-
sent the population, hence, the parametrical test based on
the mean are not applicable, instead, we have to use
a non-parametric test based on the median like the
Chebyshev Theorem and the MannWhitneysUtest.
Analysing the data, and how it varies around the
mean, we can affirm according to the Chebyshevstheo-
rem that the data from the Canon camera and the multi-
spectral sensor Sequoia fit in a free distribution of 95%
and a kvalue of 3.16, limiting the distribution ktimes the
standard deviation above and below the mean.
Considering the similarity test, or the Mann
WhitneysUtest, we calculated the Uvalues for the
pairs of values obtained from the GCP and values
obtained in the 3D model and the orthomosaic (E,
Nand hindividually) from both camera/sensors, as
Figure 11. Spatial distribution and semivariograms of the discrepancies EN (0°, 45°, 45°, 60° and 90°) where is possible to verify
that there is no correlation between the points and the distance.
presented in Tables 4 and 5for a confidence level of
95% and critical Zof 1.96.
For the MannWhitneysUtest as explained in the
prior section the null hypothesis (or H
) attest that
the two sets of data are similar, and the alternative
hypothesis is that the two set of data are different,
while the test accepts or rejects H
As we can observe in Table 4, we cannot reject the null
hypothesis only for the Eposition, as the null hypothesis
is rejected for the positions Nand h,wecaninferthatthe
positional data presents significant difference with the
data collected with the differential GNSS RTK.
In the opposite, the results for the Sequoia sensor
(Table 5) do not reject the null hypothesis implying
that the data do not have a significant difference to the
data collected in the field also indicating a better result
According to the statistical tests applied to verify the
EN and hpositional accuracy, the tests showed that
the sample satisfied the requirements of the basic
hypothesis of independence and randomness but
not for normality forcing us to apply non-
parametric tests like the MannWhitneysUtest
and the Chebyshevs theorem.
Comparing the two sensors/cameras, we found
betters results for the Sequoia sensor in comparison
against the Canon Powershot used in this work, as
the 2D (orthomosaic) and 3D (DEM/DSM) products
had lower standard deviation indicating more
precision. Although we found certain individual
points with lower discrepancy for the Canon camera
(more accuracy), the data obtained from all points
varied significantly showing no consistency, rein-
forced by the MannWhitneysUtest.
Also for data generated from both cameras/sen-
sors, we had similar outliers, where we suspect were
caused by a shadowed area given by a taller building
near these specific control points. In addition, we
believe we had valid data, considering the objective
of analysing the accuracy of the 2D and 3D models
around a water body, due to its reflective and refrac-
tive nature making the use of the algorithms during
the photogrammetry process difficult.
Comparing similar works (Fonstad et al., 2013;
Harwin & Lucieer, 2012; Jaud et al., 2016; Laliberte
et al., 2010; Lobnig et al., 2015; Neto et al., 2017;
Verhoeven, 2011), we had lower accuracy and preci-
sion, however, is not possible to affirm we had worse
or better results due differences in methodology and
the particular presence of a water body in our work.
Moreover, in order to compare similar works, we
reinforce the importance of a common protocol for
image capture and UAV flight plan, allowing the
direct comparison between.
The SfM + MVS technique is gaining importance due
to its accessibility and feasibility to mapping areas
where geometric accuracy is required, although
water bodies are excluded for assessment in the
majority of works that apply photogrammetry.
Table 3. Descriptive statistical measures of the discrepancies for the altimetry, 2D positional and 3D
positional for Sequoia and Canon sensors.
Discrepancy Mean Median Standard deviation RMSE CV
EN (m)
2D positional
Sequoia 0.330 0.258 0.182 0.375 0.552
Canon 0.386 0.256 0.345 0.514 0.894
Sequoia 0.432 0.518 0.246 0.494 0.569
Canon 0.481 0.282 0.970 1.060 2.016
DSM (m)
3D positional
Sequoia 0.544 0.579 0.306 0.620 0.793
Canon 0.617 0.381 1.030 1.178 1.669
Table 4. MannWhitneysUtest results applied to the orthomosaic and the DSM from the Canon camera.
Coord. n
E28 28 796 797 393 0.124 Dont rejects H
N28 28 793 848 342 6.186 Rejects H
H28 28 833 688 357 4.330 Rejects H
Table 5. MannWhitneysUtest results applied to the orthomosaic and the DSM from the Sequoia sensor.
Coord. n
E28 28 795 801 389 0.371 Dont rejects H
N28 28 800 796 390 0.247 Dont rejects H
h28 28 755 802 393 0.495 Dont rejects H
12 D. M. D. C. VITTI ET AL.
Two types of camera/sensors were used in this
work: the 16 MP Canon Powershot camera; and the
16 MP multispectral sensor Sequoia embedded in
a UAV hex-rotor capturing 83 images each in one
This work aimed to evaluate the accuracy of 2D
(orthomosaic) and 3D models using GCPs obtained
in the margin of an artificial lake. The photogram-
metry community less explores the presence of
a water body, and we found considerable results in
accuracy and precision for both cameras/sensors,
with the multispectral Sequoia sensor showing better
results than our modified Canon camera.
Due the nature of the overall data studied, along-
side the descriptive statistics we applied non-
parametric tests that consider free distribution, as
the MannWhitneysUtest to compare the simila-
rities between the GCP data and the data obtained
from the models, and Chebyshevs theorem to affirm
that 95% of our data fit between 3.16 times of the
standard deviation.
With the results found in this work and its statis-
tical validation, we consider that we had relevant
results analysing the data obtained from GCPs
around a water body, allowing the community to
make similar comparisons in the future.
Disclosure statement
No potential conflict of interest was reported by the
Ademir Marques Junior
Ariza, F.J., & Atkinson, A.D. (2008). Analysis of some
positional accuracy assessment methodologies. Journal
of Surveying Engineering.,4554. doi:10.1061/(ASCE)
ASPRS. (2015). Positional accuracy standards for digital
geospatial data. Photogrammetric Engineering & Remote
Atkinson, A.D. (2005). Control de Calidad Posicional en
Cartografía: Analisis dos Principales Estandares
y propuesta de mejora. (Doctoral Thesis). Universidad
de Jaen.
Brazil. (1984). Normas Técnicas da Cartografia Nacional.
Decreto Nº 89.817 de 20 de junho de 1984. Rio de Janeiro:
Fundação Instituto Brasileiro de Geográfica e Estatística.
Bustos, O. (1981). Estimação Robusta no Modelo de
Posição. Rio de Janeiro: Instituto de Matemática Pura
e Aplicada.
Debevec, P.C., Taylor, C., & MaliK, J. (1990). Modeling and
rendering architecture from department of defense.
(M. S. 60001., Ed.) Mapping, charting and geodesy accu-
racy. 136 327330 doi: 10.1099/00221287-136-2-327
Federal Geographic Data Committee. (1998). Geospatial
positioning accuracy standards part 3: National standard
for spatial data accuracy. Reston, VA: Subcommittee for
Base Cartographic Data.
Fonstad, M.A., Dietrich, J.T., Couville, B.C., Jensen, J.L., &
Carbonneau, P.E. (2013). Topographic structure from
motion: A new development in photogrammetric
measurement. Earth Surface Processes and Landforms,
38(4), 421430. doi:10.1002/esp.3366
Furukawa, Y., & Curless, B.S. (2010). Towards Internet-
scale multi view stereo. In Computer Vision and Pattern
Recognition (CVPR) (pp. 14341441). San Francisco,
Harwin, S., & Lucieer, A. (2012). Assessing the accuraccy of
georeferenced point clouds produced via multi-view
stereopsis from Unmanned Aerial Vehicle (UAV)
imagery. Remote Sensing,4(6), 15731599. doi:10.3390/
Jaud, M., Passot, S., Le Bivic, R., Delacourt, C.,
Grandjean, P., & Le Dantec, N. (2016). Assessing the
accuracy of high resolution digital surface models com-
puted by photoscan® and MicMac® in sub-optimal survey
conditions. Remote Sensing,8(6), 465. doi:10.3390/
Joaquim Höhle, M.H. (2009). Accuracy assessment of digi-
tal elevation models by means of robust statistical
methods. ISPRS Journal of Photogrammetry and Remote
Sensing,64, 398406. doi:10.1016/j.isprsjprs.2009.02.003
Kraus, K. (2004). Photogrammetrie (Vol. 1). Berlin: Walter
de Gruyter.
Laliberte, A.S., Herrick, J.E., Rango, E., & Winters, C. (2010).
Acquisition, orthorectification, and object-based classfica-
tion of Unmanned Aerial Vehicle (UAV) imagery for range-
land monitoring. Photogrammetric Engineering & Remote
Sensing,76(6), 661672. doi:10.14358/PERS.76.6.661
Lerma, J.L. (2002). Fotogrametría Moderna: Analítica
y Digital. Valencia: Universitat Politécnica de Valência.
Lobnig, K., Tscharf, A., & Mayer, G. (2015). Einfluss der
Georeferenzierung auf die absolute
Rekonstruktionsgenauigkeit der photogrammetrischen
UAV-gestützten Tagebauvermessung. (influence of geo-
referencing on the absolute accuracy of UAV-based
photogrammetric reconstructions in open pit surveying).
BHM Berg-und Hüttenmännische Monatshefte,160(8),
373378. doi:10.1007/s00501-015-0388-8
Mann, P.S. (2008). Introdução a Estatística (5 ed.).
Translated Rio de Janeiro: LTC.
Neto, F., Gripp, F.D., Jr, Botelho, J., Santos, M.F.,
Nascimento, A.P., & Fonseca, A.L. (2017). Assessment
of positional quality in spatial data generated by VANT
using point and linear feature for cadastre applications.
Boletim de Ciências Geodésicas,23(1), 134149.
Oliveira, M. (2002). Aage-based modeling and rendering
techniques: a survey. Rita,IX(2), 3766.
Preciado, A.D. (2000). Investigation sobre los Metodos de
Estimacion Robusta aplicados a problemas fundamentals
de la Fotogrametria. (Doctoral Thesis). Universidad de
Remondino, F.B., Scaioni, M., & Sarazzi, D. (2011). UAV
photogrammetry for mapping and 3D modeling cur-
rent status and future perspectives. ISPRS ICWG I/V
UAV-g (unmanned aerial vehicle in geomatics) confer-
ence. Zurich, Switzerland.
Santos, A.D., Medeiros, N.D., Santos, G.R., & Lisboa
Filho, J. (2017). Use of geostatistics on absolute posi-
tional accuracy assessment of geospatial data. Boletim de
Ciências Geodésicas,23(3), 405418. doi:10.1590/s1982-
Santos, P.A. (2015). Controle de qualidade cartográfica:
metodologias para avaliação da acurácia posicional em
dados espaciais. Doctoral Thesis, Viçosa, Minas Gerais,
Brazil: Universidade Federal de Viçosa.
Smith, M., Carrivick, J., & Quincey, D. (2016). Structure
from motion in physical geography. Progress in Physical
Geography,40, 247275. doi:10.1177/0309133315615805
Snavely, N., Seitz, S., & Szeliski, R. (2008). Modelling the
world from Internet photo collections. International
Journal of Computer Vision,80, 189210. doi:10.1007/
Soares, A. (2000). Geoestatística para as Ciências da Terra
e do Ambiente. Lisboa: Instituto Superior Técnico.
Torman, V.B., Coster, R., & Riboldil, J. (2012).
Normalidade de variáveis: Métodos de verificação
e comparação de alguns testes não-paramétricos por
simulação. Revista HCPA,32(2), 227234.
Turner, D., Lucieer, A., & Christopher Watson, C.
(2012). An automated technique for generating geor-
ectified mosaics from ultra-high resolution
Unmanned Aerial Vehicle (UAV) Imagery, based
on Structure from Motion (SfM) point clouds.
Remote Sensing,4(5), 13921410. doi:10.3390/
Verhoeven, G. (2011). Taking computer vision aloft
archaeological three dimensional reconstructions
from aerial photographs with PhotoScan.
Archaeological Prospection,18(1), 6773.
Yamamoto, J.K., & Landim, P.M. (2013). Geoestatística:
Conceitos e Aplicações. São Paulo: Oficina de Textos.
14 D. M. D. C. VITTI ET AL.
... However, in most works, the statistical methods used in the evaluation are not detailed, where it is assumed that this data is obtained from the processing software. However, this evaluation should consider parametric and non-parametric tests (Mann, 2007) which is often not done (da Silva, Souto, Duarte, Bicho, & Sabadia, 2015;Neto et al., 2017;Turner et al., 2012;Vitti et al., 2019). ...
... If the preliminary hypothesis is confirmed, it is followed by the calculations of the statistical measures and the realization of inferences employing parametric tests, otherwise, non-parametric tests are adopted. Figure 5 illustrates the protocol employed in this work, that is based on the works of Atkinson (2005), Domingo (2001), and Santos, Medeiros, Santos, and Rodrigues (2015) and that was prior established by Vitti et al. (2019). ...
... Excluding outliers using the expected error (σ priori ) To accurate estimate errors (discrepancies) all variables should be considered (acquisition and processing) (American Society Photogrammetry and Remote Sensing [ASPRS], 2015). Thus, the expected error σ priori can be estimated by Equation 1 (Vitti et al., 2019): ...
Full-text available
Geometric accuracy is an important attribute of cartographic products and UAV photogrammetry has been gaining market in topographic mapping thanks to high spatial and temporal resolution, however, they need proper evaluation following accuracy standards and protocols. Regarding this, this work evaluates products from digital photogrammetry from images acquired with a fixed-wing UAV (18Mpixel camera) in a 300-380m height flight over a Hydroelectric Power Plant (HPP) in Brazil. A dataset of 23 ground control points assessed with an RTK-GNSS (using natural targets) was validated with its homologous in the Digital Surface Model (DSM) and the orthomosaic, following a workflow in which the appropriate statistics were applied. Following parametric tests like the Students t-test and the Chi-square, we compared the results with the Brazilian Cartographic Standard for digital cartography, achieving minimum scale of 1: 20,000 (RMSE of 1.04 m) for the orthomosaic, and minimum scale of 1: 10,000 (RMSE of 1.31 m) for the elevation in the DSM, although, no special targets were used. As the 3D mapping generated using the photogrammetry still needs a protocol to evaluate the accuracy, this work applied a proposed workflow respecting the quality of the data to meet the requirements of the cartographic standard.
... The IQR outlier removal adjusted to the skewness of the data shows how important it is compared to the usual IQR. Also, following the trend to analyze geographical data using nonparametric tests [15][13] [16] this work evaluated the models using robust statistical measures (based on the median). ...
Conference Paper
Full-text available
This work evaluates robust statistical assessments for 3D models generated by the process of photogrammetry using images captured from Google Earth screenshots in a method named Printgrammetry. This method can make 3D models more accessible to be explored in other ways, including virtual reality and use in other programs with incorporated tools that are not available in the Google Earth. To verify the quality of the Printgrammetry model we proposed the use of skewness-adjusted robust statistics, an approach that avoids the exclusion of relevant data. The results confirmed that the proposed outlier detection method is more realistic to the skewness of the data than the usual outlier detection methods. The robust statistical methods adequate to our data assessed the quality of the Printgrammetry model.
... The Parrot Sequoia+ multispectral camera is mainly used in precision agriculture, and there are very few publications concerning its utilization for hydrological investigations. Previous studies referred to, among others, SSCs in different water layers in rivers [58], the precision of digital surface models (DSM) in water pools based on multispectral camera pictures [67], or assessments of the ecological integrity of wetlands through multispectral mapping [68]. Nonetheless, there are no findings confirming the applicability of using multispectral UAVs for research on polar marine areas. ...
Full-text available
A Parrot Sequoia+ multispectral camera on a Parrot Bluegrass drone registered in four spectral bands (green, red, red edge (RE), and near-infrared (NIR)) to identify glacial outflow zones and determined the meltwater turbidity values in waters in front of the following Antarctic glaciers: Ecology, Dera Icefall, Zalewski, and Krak on King George Island, Southern Shetlands was used. This process was supported by a Red-Green-Blue (RGB) colour model from a Zenmuse X5 camera on an Inspire 2 quadcopter drone. Additional surface water turbidity measurements were carried out using a Yellow Springs Instruments (YSI) sonde EXO2. From this research, it was apparent that for mapping low-turbidity and medium-turbidity waters (<70 formazinenephelometricunits (FNU)), a red spectral band should be used, since it is insensitive to possible surface ice phenomena and registers the presence of both red and white sediments. High-turbidity plumes with elevated FNU values should be identified through the NIR band. Strong correlation coefficients between the reflectance at particular bands and FNU readings (RGreen = 0.85, RRed = 0.85, REdge = 0.84, and RNIR = 0.83) are shown that multispectral mapping using Unmanned Aerial Vehicles (UAVs) can be successfully usedeven in the unfavourable weather conditions and harsh climate of Antarctica. Lastly, the movement of water masses in Admiralty Bay is briefly discussed and supported by the results from EXO2 measurements.
Conference Paper
Hand Gesture Recognition (HGR) is a system to translate the hand gestures express to literature, which is a natural way of communication between deaf-mutes and non-disabled people. However, due to the complexity of relative positions of fingers, hands sizes, and environmental illumination, the hand gesture recognition is difficult. In this paper, a Fully Connected Neural Network (FCNN) algorithm for RGB-D sensor based HGR is proposed. We firstly build datasets of fingers joints and the center coordinates of hands in 3 dimensions. Then we normalize the samples to eliminate the natural difference of hands. Finally, the data are classified using a 3 layers FCNN. Totally 13,000 data of 26 hand gestures are collected. We randomly select 80% of these data for training and 20% of them for testing. According to the experiments, the average recognition accuracy is 94.73%.
Full-text available
The paper presents an efficient methodology of water body extent estimation based on remotely sensed data collected with UAV (Unmanned Aerial Vehicle). The methodology includes the data collection with selected sensors and processing of remotely sensed data to obtain accurate geospatial products that are finally used to estimate water body extent. Three sensors were investigated: RGB (Red Green Blue) camera, thermal infrared camera, and laser scanner. The platform used to carry each of these sensors was an Aibot X6—a multirotor type of UAV. Test data was collected at 6 sites containing different types of water bodies, including 4 river sections, an old river bed, and a part of a lake shore. The processing of collected data resulted in 2.5-D and 2-D geospatial products that were used subsequently for water body extent estimation. Depending on the type of used sensor, the created geospatial product, and the type of the water body and the land cover, three strategies employing image processing tools were developed to estimate water body range. The obtained results were assessed in terms of classification accuracy (distinguishing the water body from the land) and geometrical planar accuracy of the water body extent. The product identified as the most suitable in water body detection was four bands RGB+TIR (Thermal InfraRed) ortho mosaic. It allowed to achieve the average kappa coefficient of the water body identification above 0.9. The planar accuracy of water body extent varied depending on the type of the sensor, the geospatial product, and the test site conditions, but it was comparable with results obtained in similar studies.
Full-text available
In the area of Geosciences it is intuitive to think of spatial correlation as a phenomenon under study, and Geostatistics has tools to identify and represent the behavior of such dependency. The spatial analysis of the results of an inspection of the quality of a cartographic product is generally not addressed in the standards, which are restricted to descriptive and tabular findings, based on the assumption of the Classical Statistics of independence of observed data. At the Brazilian National Infrastructure of Spatial Data (INDE), various cartographic products should be made available to society, along with their metadata. This paper proposes a methodology for quality inspection based on international standards and on the Cartographic Accuracy Standard (PEC), using geostatistical methods and spatial representation of this benchmarking, through positional quality maps. The method of evaluating the quality of data was applied to Brazil's Continual Cartographic Base, at the scale of 1:250000 - BC250, with a focus on absolute positional accuracy. The quality map generated presented regionalizations of the planimetric error confirmed by the producer team of the referred cartographic base of IBGE. Such information can help users and producers to understand the spatial behavior of cartographic product quality in study
Full-text available
Currently the validation of the quality of cartographic information has been an issue ever more evident and important. With the emergence of the National Spatial Data Infrastructure (INDE) in Brazil, there were several technical specifications in order to regulate and standardize the many processes related to national cartographic production. In this context, the objective of this study is to evaluate the positional planimetric accuracy of orthophoto generated from an embedded sensor on a UAV platform by applying methods of point and linear features, according to the Brazilian standard for positional accuracy (Decree Law 89 817 combined with ET-ADGV), considering as a reference survey performed using RTK GNSS positioning technology. The results obtained demonstrate that both linear feature method (Double Buffer method) and the pontual feature method showed similar results, obtaining classification Class B to scale 1: 1,000, according to the Decree Law No. 89,817 / ET-ADGV
Full-text available
For monitoring purposes and in the context of geomorphological research, Unmanned Aerial Vehicles (UAV) appear to be a promising solution to provide multi-temporal Digital Surface Models (DSMs) and orthophotographs. There are a variety of photogrammetric software tools available for UAV-based data. The objective of this study is to investigate the level of accuracy that can be achieved using two of these software tools: Agisoft PhotoScan® Pro and an open-source alternative, IGN© MicMac®, in sub-optimal survey conditions (rugged terrain, with a large variety of morphological features covering a range of roughness sizes, poor GPS reception). A set of UAV images has been taken by a hexacopter drone above the Rivière des Remparts, a river on Reunion Island. This site was chosen for its challenging survey conditions: the topography of the study area (i) involved constraints on the flight plan; (ii) implied errors on some GPS measurements; (iii) prevented an optimal distribution of the Ground Control Points (GCPs) and; (iv) was very complex to reconstruct. Several image processing tests are performed with different scenarios in order to analyze the sensitivity of each software package to different parameters (image quality, numbers of GCPs, etc.). When computing the horizontal and vertical errors within a control region on a set of ground reference targets, both methods provide rather similar results. A precision up to 3–4 cm is achievable with these software packages. The DSM quality is also assessed over the entire study area comparing PhotoScan DSM and MicMac DSM with a Terrestrial Laser Scanner (TLS) point cloud. PhotoScan and MicMac DSM are also compared at the scale of particular features. Both software packages provide satisfying results: PhotoScan is more straightforward to use but its source code is not open; MicMac is recommended for experimented users as it is more flexible.
Full-text available
The use of unmanned aerial vehicles (UAVs) for natural resource applications has increased considerably in recent years due to their greater availability, the miniaturization of sensors, and the ability to deploy a UAV relatively quickly and repeatedly at low altitudes. We examine in this paper the potential of using a small UAV for rangeland inventory, assessment and monitoring. Imagery with a ground resolved distance of 8 cm was acquired over a 290 ha site in southwestern Idaho. We developed a semi-automated orthorectification procedure suitable for handling large numbers of small-footprint UAV images. The geometric accuracy of the orthorectified image mosaics ranged from 1.5 m to 2 m. We used object-based hierarchical image analysis to classify imagery of plots measured concurrently on the ground using standard rangeland monitoring procedures. Correlations between image-and ground-based estimates of percent cover resulted in r-squared values ranging from 0.86 to 0.98. Time estimates indicated a greater efficiency for the image-based method compared to ground measurements. The overall classification accuracies for the two image mosaics were 83 percent and 88 percent. Even under the current limitations of operating a UAV in the National Airspace, the results of this study show that UAVs can be used successfully to obtain imagery for rangeland monitoring, and that the remote sensing approach can either complement or replace some ground-based measurements. We discuss details of the UAV mission, image processing and analysis, and accuracy assessment.
Full-text available
Unmanned Aerial Vehicles (UAVs) are an exciting new remote sensing tool capable of acquiring high resolution spatial data. Remote sensing with UAVs has the potential to provide imagery at an unprecedented spatial and temporal resolution. The small footprint of UAV imagery, however, makes it necessary to develop automated techniques to geometrically rectify and mosaic the imagery such that larger areas can be monitored. In this paper, we present a technique for geometric correction and mosaicking of UAV photography using feature matching and Structure from Motion (SfM) photogrammetric techniques. Images are processed to create three dimensional point clouds, initially in an arbitrary model space. The point clouds are transformed into a real-world coordinate system using either a direct georeferencing technique that uses estimated camera positions or via a Ground Control Point (GCP) technique that uses automatically identified GCPs within the point cloud. The point cloud is then used to generate a Digital Terrain Model (DTM) required for rectification of the images. Subsequent georeferenced images are then joined together to form a mosaic of the study area. The absolute spatial accuracy of the direct technique was found to be 65–120 cm whilst the GCP technique achieves an accuracy of approximately 10–15 cm.
Accurate, precise and rapid acquisition of topographic data is fundamental to many sub-disciplines of physical geography. Technological developments over the past few decades have made fully distributed data sets of centimetric resolution and accuracy commonplace, yet the emergence of Structure from Motion (SfM) with Multi-View Stereo (MVS) in recent years has revolutionised three-dimensional topographic surveys in physical geography by democratising data collection and processing. SfM-MVS originates from the fields of computer vision and photogrammetry, requires minimal expensive equipment or specialist expertise and, under certain conditions, can produce point clouds of comparable quality to existing survey methods (e.g. Terrestrial Laser Scanning). Consequently, applications of SfM-MVS in physical geography have multiplied rapidly. There are many practical options available to physical geographers when planning a SfM-MVS survey (e.g. platforms, cameras, software), yet, many SfM-MVS end-users are uncertain as to the errors associated with each choice and, perhaps most fundamentally, the processes actually taking place as part of the SfM-MVS workflow. This paper details the typical workflow applied by SfM-MVS software packages, reviews practical details of implementing SfM-MVS, combines existing validation studies to assess practically achievable data quality and reviews the range of applications of SfM-MVS in physical geography. The flexibility of the SfM-MVS approach complicates attempts to validate SfM-MVS robustly as each individual validation study will use a different approach (e.g. platform, camera, georeferencing method, etc.). We highlight the need for greater transparency in SfM-MVS processing and enhanced ability to adjust parameters that determine survey quality. Looking forwards, future prospects of SfM-MVS in physical geography are identified through discussion of more recent developments in the fields of image analysis and computer vision.
Caused by increasing use of unmanned aerial systems in the various fields of surveying, the question of accuracy has been growing in importance. One of the main accuracy influencing factors is certainly the geo-referencing. Due to the fact that purely image-based reconstruction methods are not able to generate metrically correct results in a defined coordinate frame, the step of referencing has to be carried out regardless of the images captured. For this reason, a bachelor thesis concerning the influence of geo-referencing on the accuracy of photogrammetric reconstructions was carried out at the Chair of Mining Engineering and Mineral Economics at Montanuniversitaet Leoben. Specifically, the influence of the spatial distribution of control points was tested in form of a point by point analysis based on several independent aerial surveys of a typical wall at the Styrian Erzberg. The results clearly demonstrate the systematic relationship between the control point assembly and the achievable accuracy. To ensure defined accuracy in the entire survey area, a uniform spatial distribution of the control points in the whole scene is essential.
The production of topographic datasets is of increasing interest and application throughout the geomorphic sciences, and river science is no exception. Consequently, a wide range of topographic measurement methods have evolved. Despite the range of available methods, the production of high resolution, high quality digital elevation models (DEMs) requires a significant investment in personnel time, hardware and/or software. However, image-based methods such as digital photogrammetry have been decreasing in costs. Developed for the purpose of rapid, inexpensive and easy three-dimensional surveys of buildings or small objects, the ‘structure from motion’ photogrammetric approach (SfM) is an image-based method which could deliver a methodological leap if transferred to geomorphic applications, requires little training and is extremely inexpensive. Using an online SfM program, we created high-resolution digital elevation models of a river environment from ordinary photographs produced from a workflow that takes advantage of free and open source software. This process reconstructs real world scenes from SfM algorithms based on the derived positions of the photographs in three-dimensional space. The basic product of the SfM process is a point cloud of identifiable features present in the input photographs. This point cloud can be georeferenced from a small number of ground control points collected in the field or from measurements of camera positions at the time of image acquisition. The georeferenced point cloud can then be used to create a variety of digital elevation products. We examine the applicability of SfM in the Pedernales River in Texas (USA), where several hundred images taken from a hand-held helikite are used to produce DEMs of the fluvial topographic environment. This test shows that SfM and low-altitude platforms can produce point clouds with point densities comparable with airborne LiDAR, with horizontal and vertical precision in the centimeter range, and with very low capital and labor costs and low expertise levels. Copyright © 2012 John Wiley & Sons, Ltd.
This work presents an analysis of some standard methodologies for positional accuracy assessment of geographic data bases, taking into account aspects like the statistical formulation, the size of the control sample, the distribution and typology of the control elements, etc. Here we point out some weaknesses of the majority of standards: scarce formalism, inappropriate terminology for dealing with uncertainty, minimum recommended sample size, no assessment of the base hypothesis of the statistical model being applied, no information about the statistical behavior and reliability of the method, etc. The analysis developed here can serve as a starting point for the development of improved methodologies.