Title | Student Name GEOM7000 practical 1 4 Draft |
---|---|
Course | Introduction To Remote Sensing Of Environment |
Institution | University of Queensland |
Pages | 23 |
File Size | 1.7 MB |
File Type | |
Total Downloads | 69 |
Total Views | 134 |
Sample practical assignment 1...
Linking Field and Image Data
http://www.gps.gov/multimedia/images 2013/03/22
GPS-III Satellite. Source: Image courtesy of gps.gov
Author: GEOM 7000 Assignment 1 Lecturer: Professor S Phinn Field Work Date: Group 4
TABLE OF CONTENTS INTRODUCTION ........................................................................................................................ 2 FIELD EQUIPMENT ............................................................................................................................ 2 SOURCE IMAGES ............................................................................................................................... 2
QUESTIONS AND ANSWERS ................................................................................................ 3 TASK 1.1 GPS READINGS AT A POINT ................................................................................................... 3 DESCRIPTION ............................................................................................................................... 3 RESULTS...................................................................................................................................... 3 TASK 1.2 AND 1.3 - GPS READINGS FOR GROUND AREAS ........................................................................ 4 DESCRIPTION ............................................................................................................................... 4 RESULTS...................................................................................................................................... 5 TASK 2 - WHAT IS IN A PIXEL? ............................................................................................................. 7 DESCRIPTION ............................................................................................................................... 7 RESULTS...................................................................................................................................... 7 TASK 3 -SPECTRAL SIGNATURES ........................................................................................................ 11 DESCRIPTION ............................................................................................................................. 11 RESULTS.................................................................................................................................... 11 TASK 4 - TRANSECT ......................................................................................................................... 14 TASK DESCRIPTION ..................................................................................................................... 14 RESULTS.................................................................................................................................... 14
APPENDICES .......................................................................................................................... 17 TABLE OF FIGURES .......................................................................................................................... 17 LIST OF TABLES ............................................................................................................................... 17 FIELD DATA FILES ........................................................................................................................... 18
BIBLIOGRAPHY .................................................................................................................. 21
1
Introduction As a r
This report will assess data collected in the field at UQ St Lucia Campus, and data from remotely sensed imagery, and provide possible reasons for discrepancies between the data sources.
Field Equipment The following equipment was used to take measurements in the field:
Garmin eTrex Legend® H handheld GPS unit
50 m Tape
ASD HandHeld 2 Portable Spectrometer
FPC measurement tool.
Source Images The images that were used for comparison to field data are described in the tables below.
All Images
Datum
Projection
Units
WGS-84
UTM Zone 56S
Metres
Table 1: Datum, Projection and Units
Image
Pixel size (m)
Bands
File date
Collection date
Upper Left
StLucia_Orthophoto
Approx. 0.15
3
14/9/12
2009
500159.5500 E
2.4
4
(Orthophoto) StLucia_Quickbird
6959689.8000 N 20/9/12
?
(Quickbird) StLucia_Landsat_TM (Landsat) Table 2: Image Details
2
500158.4250 E 6959690.9250 N
25
7
14/9/12
17/1/2011
474122.5000 E 6970147.5000 N
QUESTIONS AND ANSWERS Task 1.1 GPS readings at a point Description The GPS unit was placed on the ground and waypoints 001 to 006 were recorded at one-minute intervals. The GPS unit averaging function was applied for waypoint 006.
Results The results are illustrated in Figure 1 below.
[Image] b Field Data Collection - Task 1.1 (From Left)
a Plot of GPS Coordinates
c GPS Coordinates overlaid on Orthophoto Figure 1: GPS readings of the same physical point
3
d Orthophoto at 4 times zoom
Why are the readings different? The readings are different because the uent satellite signals to adjust the po reading for errors in rangi he time and locati cast by the satellite, the speed nd the elapsed time from t. (National Coordination Office for Space-Based Positioning, 2013) The calculation is susceptible to a variety of factors including: and the GPS unit clocks; var ob
) (Lillesand et al., 2008).
As a form of electromagnetic radiation, the signal can be absorbed, transmitted or reflected by vegetation (Phinn, 2013). GPS signal disruption also increases with (Pirti, 2005). The diagram in Figure 1a (above) illustrates that the GPS readings are generally ‘moving’ direction; while the images in Figure ow that there is more variance in the Northing measurement py may be to the north or north-east of the location. Any signal obstruction or attenuation caused by the ca
(Jensen and Jensen, 2013).This results in different GPS readings at different points in time.
Table 3: GPS Waypoint Data and Statistics
Task 1.2 and 1.3 - GPS readings for ground areas Description The GPS unit was placed on the ground at points on the circumference of each feature. Waypoints 007 to 016 were recorded for the man-made feature and the natural feature. The GPS unit averaging function was used for waypoint 007 to 012.
4
Results The results are illustrated in Figure 2 below.
a Plot of GPS Coordinates
b GPS Coordinates overlaid on Orthophoto
c GPS Coordinates overlaid on Quickbird image at 13 times zoom Figure 2: GPS readings of ground areas
5
Explain why your measured locations may not match up to where the corresponding feature is located in the St Lucia Orthophoto? The locations may not match up to the corresponding feature due to e n accuracy. An orthophoto is processed to remove distortions c Lillesand et al., 2008), w (Phinn, 2013, Jensen and Jensen, 2013). The photograph must also be georeferenced to a datum, coordinate system and projection, wher ents a three the grid referencing system and the projection represents equations to two (Phinn, 2013). Some inaccuracy may result from projection, for example, distortions i (Jensen and Jensen, 2013).
Root Me any errors in processing,
e points on the photograph and the map (Jensen, 2005). In addition to
There are several sources of GPS reading error as discussed in Task 1.1 above. The accuracy of
6
Task 2 - What is in a pixel? Description A waypoint was taken for the upper North West point (the ‘Origin’) of a four sided polygon, representing the upper left corner of a pixel. A tape measure was used to measure distances, and subsequent waypoints were taken at each corner of the polygon. Measurements and waypoints were then taken for the next size pixel. Note that the same origin coordinates are used for all pixel sizes.
Results The field results are summarised in Table 4 and illustrated on the Quickbird image in Figure 3 (below).
Pixel size (metres) 0.5 x 05
2.4 x 2.4
10 x 10
20 x 20
30 x 30
50 x 50
Corner GPS Coordinates
% Ground Cover Type
Figure
500965.26 E, 6959029.32 N (Origin) 500965.72 E, 6959029.78 N 500965.72 E, 6959029.05 N 500965.84 E, 6959028.85 N Origin 500969.64 E, 6959030.29 N 500970.06 E, 6959027.76 N 500968.45 E, 6959028.08 N Origin 500978.53 E, 6959032.36 N 500976.51 E, 6959020.22 N 500965.58 E, 6959019.48 N Origin 500984.39 E, 6959030.95 N 500984.94 E, 6959011.47 N 500970.76 E, 6959010.88 N Origin 500995.23 E, 6959033.13 N 500996.75 E, 6958993.43 N 500967.86 E, 6958999.42 N Origin 501014.63 E, 6959033.94 N 501014.64 E, 6958973.02 N 500972.09 E, 6958980.87 N
100% Leaf litter
Refer to Figure 4b (below)
100% Leaf litter
Refer to Figure 4b (below)
90% Leaf litter 10% Small trees
Refer to Figure 4a (below)
80% Leaf litter 10% Small trees 10% Grass
Refer to Figure 4a (below)
5% Leaf litter 45% Small trees 55% Grass
Refer to Figure 4a (below)
5% Leaf litter 15% Small trees 80% Grass
Refer to Figure 4a (below)
Table 4: Summary of pixel measurement results
7
b Orthophoto at 5 times zoom to show 0.5m pixel coordinates
a Pixels drawn on Orthophoto Figure 3: Pixel coordinates overlaid on Orthophoto
Explain how the size of the image pixel relates to the size of features able to be identified on an image. The image pixel needs to be smaller than the features to be identified, but other factors such as contrast, (Campbell, 2002), Jensen (2007) suggests that
As illustrated in Figure 4 (below), the higher spatial resolution of the Orthophoto (Figure 4a) provides the ability to identify e Quickbird image (Figure 4b) we can still identify
8
Visible features: Individual cars vegetation, some individual trees, some road markings, footpaths, features less than 2m2.
a Orthophoto - 0.15 m pixel (approx.)
Visible features: Sports fields, individual houses, roads, cars, parking areas.
b Quickbird image - 2.4 m pixel
Visible features: Large features e.g. river; main roads; groups of similar colours e.g.: artificial surfaces, vegetation patches, features wider than 20m and longer than 50m.
c Landsat image - 25 m pixel Figure 4 : Visible features by Pixel Size
9
Measurements from the Orthophoto are illustrated in Figure 5 (below), and Table 5 provides a comparison of visibility with respect to the Quickbird image (Figure 4b).Table 5: Visibility - Feature Size vs. Pixel Size
b Drain cover approximately 1.9 m wide
a Tree crown approximately 10 m across
c Tennis Court approximately 11 m across Figure 5: Measurement of smallest dimension from Orthophoto
Feature
Measurement m(approximate)
Ability to identify feature Orthophoto image (0.15 m pixel)
Quickbird image (2.4m pixel)
Tree
10m
High
Low
Drain cover
1.9m
High
No (Not visible)
Tennis Court
10m (Length approx. 24m)
High
Medium
Table 5: Visibility - Feature Size vs. Pixel Size
sh feature size.
10
Based on the above comparison, the appropriate relationship for pixel size is
Task 3 -Spectral Signatures Description The Spectrometer was used to measure the reflectance of Spectralon as the diffuse calibration target and then various objects were measured.
Results The spectral signatures and objects measured are illustrated below in Figure 6 and Figure 7 respectively.
Figure 6: In situ Spectral Signatures
11
a: Grass - short
b: Hedge
c: Leaves - Yellow
d: Grass - dead
e: Leaves - Dead
f: Leaf litter
g: Water - Black Bucket
h: Water - Red Bucket
j: Road
k: Concrete
i Bare dirt
l: Red Brick Figure 7: Objects measured with the Field Spectrometer
12
What differences would you expect between Landsat TM Spectral Profiles and Ground Reference Data? The key differences between the Landsat TM Spectral Profiles and Ground Reference Data will be:
The impact o The impact o The different The impact o
Firstly, unlike the ground reference instrument,
e(Jensen, 2007). iation received h is influenced by factors view
l (Jensen, 2007). ce, whereas the Landsat sensor is (Phinn, 2013). Secondly, as illustrated in ‘Task 2 - What is in a pixel?’ Figure 3 (above), g (Jensen, 2005). Thirdly, as illustrated in Figure 6 (above),
The Landsat bands and profiles are illustrated in Figure 8 (below). Finally, the Landsat images are a
time that the remotely sensed imagery is captured.
Which Band from LANDSAT TM will be best for mapping basic cover types?
As illustrated in Figure 8 (below), Bands
provide good differentiation between basic types
demonstrates the greatest variance between the cover types.
13
Figure 8: Spectral Profiles from Landsat TM Image
Table 6: Bandwidth Standard Deviation
Task 4 - Transect Task Description A 50m distance was measured using a tape measure and a compass. Data was collected at 2m intervals. GPS coordinates were captured for the start and end points.
Results
14
a Vectors overlaid on Quickbird image
b At 4 x zoom
c Transect Profile Figure 9: Quickbird false colour composite and associated Transect Profile
15
Figure 10: FPC Data with Near Infrared and Red Band Transect Profiles
Compare and contrast FPC values against image values and explain why these should co-vary. Foliage Projective Cover (FPC) is a
16
Appendices Table of Figures Figure 1: GPS readings of the same physical point .............................................................................................3 Figure 2: GPS readings of ground areas ..............................................................................................................5 Figure 3: Pixel coordinates overlaid on Orthophoto ..........................................................................................8 Figure 4 : Visible features by Pixel Size ...............................................................................................................9 Figure 5: Measurement of smallest dimension from Orthophoto .................................................................. 10 Figure 6: In situ Spectral Signatures................................................................................................................. 11 Figure 7: Objects measured with the Field Spectrometer ............................................................................... 12 Figure 8: Spectral Profiles from Landsat TM Image ......................................................................................... 14 Figure 9: Quickbird false colour composite and associated Transect Profile .................................................. 15 Figure 10: FPC Data with Near Infrared and Red Band Transect Profiles ........................................................ 16
List of Tables Table 1: Datum, Projection and Units .................................................................................................................2 Table 2: Image Details .........................................................................................................................................2 Table 3: GPS Waypoint Data and Statistics .........................................................................................................4 Table 4: Summary of pixel measurement results ...............................................................................................7 Table 5: Visibility - Feature Size vs. Pixel Size .................................................................................................. 10 Table 6: Bandwidth Standard Deviation .......................................................................................................... 14
17
Field Data Files GPS Locations
18
What is in a Pixel?
19
Spectral Signatures
20
FPC
BIBLIOGRAPHY ARMSTRON, JOHN D, DENHAM, ROBERT J, DANAHER, TIM J, SCARTH, PETER F, MOFFIET,TREVOR N 2009. Prediction and validation of foliage projective cover from Landsat-5 TM and Landsat-7 ETM+ imagery. Journal of Applied Remote Sensing, 3. BARRETT, E. C. & CURTIS, L. F. 1999. Introduction to environmental remote sensing, Cheltenham, Glos., UK :, Stanley Thornes Publishers. BURKE, H. & SMITH, C. 2004. The archaeologist's field handbook [Online]. Crow's Nest, N.S.W. :: Allen & Unwin. Available: http://www.scribd.com/doc/40306556/The-Archaeologist-s-Field-Handbook [Accessed April 5 2013]. CAMPBELL, J. B. 2002. Introduction to Remote Sensing, New York, The Guildford Press. GARMIN. 2008. Garmin etrex Legend H Manual [Online]. Available: https://buy.garmin.com/shop/store/manual.jsp?product=010-00779-00&cID=145&pID=30120 [Accessed 28 March 2013].
21
GEOSCIENCE AUSTRALIA. 2013. Landsat [Online]. Commonwealth of Australia (Geoscience Australia). Available: http://www.auslig.gov.au/earth-observation/satellites-and-sensors/landsat.html [Accessed 07 April 2013]. JENSEN, J. R. 2005. Introductory digital image processing : a remote sensing perspective, Upper Saddle River, N.J. :, Prentice Hall. JENSEN, J. R. 2007. Remote Sensing of the Environment - An Earth Resource Perspective, Upper Saddle River, Pearson Education Inc. JENSEN, J. R. & JENSEN, R. R. 2013. Introductory Geographic Information Systems, Glenview, Pearson Education. LILLESAND, T. M., KIEFER, R. W. & CHIPMAN, J. W. 2008. Remote sensing and image interpretation, Hoboken, NJ :, John Wiley & Sons. NATIONAL COORDINATION OFFICE FOR SPACE-BASED POSITIONING, NAVIGATION, AND TIMING,. 2013. How GPS Works [Online]. National Coordination Office for Space-Based Positioning, Navigation, and Timing. Available: http://www.gps.gov/multimedia/poster/poster.txt [Accessed April 06 2013]. PHINN, S. R. 2013. GEOM7000 - Remote Sensing of the Environment. Brisbane: University of Queensland. PIRTI, A. 2005. Using GPS near the forest and quality control. Survey review - Directorate of Overseas Surveys, 38, 286-298.
22...