3. What kinds of information can we extract from
imagery data?
In case of Camera
Color
Directional Vector
Or Geometric information.
4. Principle of geometric measurement from Imagery
Position and Attitude of
Camera when taking a picture
or an image
Vector of light or ray from an object
(3D directional vector)
5. Principle of 3D measurement using Stereo Imagery
3D coordinates of an object can
be determined as an intersection
point of two light rays.
6. More robust and accurate measurement
from a series of images.
7. Mathematical formulation
i ) co-linearity equation
CRS: Coordinate System
O: Center of Projection (Focus p.)
(X0,Y0,Z0) : Ground CRS
z
κ
Sensor CRS
(x,y,z)
φ
Rotating angle of this coor.sys.
(ω,φ,κ) : 3 axis attitude
y
x ω
(地上座標系からみたセンサ座標系の回転角(傾き))
Image plane (Film plane)
parallel to xy plane of sensor CRS
f: focal length
Z
X
Ai (Xa,Ya,-f): Image point of A
Based on sensor CRS
Ray
A
Y
(X1, Y1,Z1)
Ground CRS
Co-linearity Equation(共線条件式)
O, Ai, A are on the same ray (straight line) in 3D space
8. Co-linearity Equation(共線条件式)
a11(X1-X0)+a21(Y1-Y0)+a31(Z1-Z0)
X = f a13(X1-X0)+a23(Y1-Y0)+a33(Z1-Z0)
Y = f a12(X1-X0)+a22(Y1-Y0)+a32(Z1-Z0)
a13(X1-X0)+a23(Y1-Y0)+a33(Z1-Z0)
Image coordinate
of a target
Ground coordinate of
a target
Sensor position
aij = aij(ω,φ,κ): Rotation Matrix
Sensor attitude
9. Estimation of sensor position and attitude using GCP
(External orientation)
Position and
attitude of sensor cood.sys.
(X0, Y0,Z0) : Position
(ω,φ,κ): Attitude
six unknown
parameters
GCP’s image coordinates
Ai(xa, ya, -f)
A(X1,Y1,Z1)
Given : f (focal length)
GCP: Ground Control Point
Collinearity Eq.
..............
xa = f ..............
..............
ya = f ..............
Non-linear least squares
method
Estimated
^ ^ ^
(X0, Y0, Z0)
^ ^
(ω,φ,κ) ^
10. 3D measurement with stereo images
Imaging plane
Image
coordinates have
to be measured
Image or sensor with
given or estimated
position/attitude
(x,y,z)
Image or sensor with
given or estimated
position/attitude
11. Basic Concept for 3D Building Extraction
r
G
r
o
u
n
o
o
f
d
3D information is key to differentiate the roofs
from the objects on the ground
12. TLS (Three Line Scanner);
Example of Digital Camera
for 3D Mapping
Stabilizer
データ処理
装置
Gyro
3 line CCD
array
GPS
位置デー
タ
進行
方向
画像表示
装置
データ記録
装置
後方
鉛直
前方
■Specifications
・Resolution 10cm(x-y)、20cm(z)
・continuous strip of digital imagery
・B/W and color imaging
13. Imaging mode of TLS
■Stereo (triplet) images can
be acquired simultaneously
Aft image
Nadir image
Fore image
14. Images taken from different angles
• It can acquire the images from three different view point.
• No distortion of altitude comparison in flight direction.
Forward
2014/2/17
Nadir
Backward
14
22. b) Synthetic Aperture Radar
- Applying pulse compression for along track (azimuth)
direction
Ground resolution : 1m~
Improving ground resolution by using Doppler effect
23.
24.
25.
26.
27.
28.
29. b) Change in frequency of return
signal due to Doppler effect
c) Characteristic of matched filter
d) Output from matched-filter
for receiving point target A
30. Geometry of Radar Image
sensor
incident
wave
angle of incidence
aspect angle
surface
direction
of flight
angle of incidence
off-nadir angle
azimuth direction
Distortions of Radar Imagery
B' A'
AB
range
direction
34. Principles of Remote Sensing
Acquiring information of objects through electromagnetic wave
reflected or radiated by the objects
Platform
Sun
Sensor
Atmosphere
Radiation
Spectral reflection
Strengths:
object
Simultaneous observation of
wide areas
Homogeneous data
Digital data
Limitations, problems:
Reference data is required for quantitative measurement
Only information reflected in electromagnetic wave
can be observed.
(Only "visible" objects!)
41. Spectral reflectance of vegetation, soil and water:
(By measuring reflectance of each spectrum, objects can be identified.)
42. Spectral reflectance of tree species:
(By measuring reflectance of each spectrum, objects can be identified.)
43. Spectral reflectance of rocks and minerals:
(By measuring reflectance of each spectrum, objects can be identified.)
44. Physical features that could be measured with
electromagnetic wave
Wave length
U.V.
Visible
I.R.
0.1 micro meter
(100nm)
1.0 micro meter
Ozone hole
Land cover/use
Vegetation (primary production)
10.0 micro meter
100. micro meter
Ground surface temperature
Sea surface temperature
1mm
Microwave
1cm
10cm
100cm
Precipitation
Sea surface wind (direction, velocity)
Wave height, direction
Snow depth
Soil water content
vegetation biomass
(standing biomass)
48. Measurement model; how to associate
sensor data with physical properties
data
The other
environmental model
affecting
Environmental model
in a broader sense
Sensor model (sensitivity)
Atmospheric model
Active
sensor
affecting
Sun
(Passive sensor)
affecting
Platform model
(fluctuation in position/attitude)
affecting
Electromagnetic wave model
(propagation, absorption, scattering…)
affecting
Object model
Radiation/reflection
Shape/geometry
Seasonal change/movement etc.
Estimating “truth” with limited
observation data with MLE or
Maximum Likelihood Estimation.
(最尤推定)
50. 1) Sensor Types for Remote Sensing
Sensors
Passive
Non Scanning
Type....Cameras
Scanning Type...
(Scanners)
Active
Non Scanning
Type..
Scanning Type...
E
X
A
M
P
L
E
S
- Photogrametric CCD Image Sensors
camera
Multispectral Scanners
- Multispectral
Microwave Radiometer
camera
-Sea surface temperature,
Vapor content,
Salt content of water etc.
- Total Station
(Range
Measurement)
- LIDAR
- Microwave
altimeter
- Geoid, Sea
surface height etc.
Microwave
scatterometer
- Velocity and direction of
sea surface wind
- Intensity of rainfall
- Water content of soil etc.
Imaging radar
- Synthetic Aperture Radar
- Side Looking Radar
(Real Aperture Radar)
Laser Range Imager
55. NOAA AVHRR Data received at AIT
-Data receiving started from Oct. 1997.
-Improvement of Processing software is on-going (by
Aug.).
-geometric correction
(extending GCP files to SE Asia)
-atmospheric correction
-Processed data delivery may start from Sept.(personal
anticipation)
62. trees
Hyper-spectral Sensors
Yasuda Hall
Gotenshita field
6000
Vegetation
Asphalt
5000
Athletic Field
Hall
4000
Pond
3000
2000
1000
0
400
Sanshiro pond
500
blue
600
green
700
800
900
1000
Near Infrared
Asphalt
(Hongo street)
(東京大学生産技術研究所 安岡研究室提供)
(単位:nm(ナノメータ))
64. 22.
0
y = 0.
0839x + 9.
7174
R ²= 0.
812
ace
S urf Tem p. 1℃)
(
20.
0
18.
0
16.
0
補正後平均値
14.
0
線形 (補正後平均値)
12.
0
10.
0
20.
0
30.
0
40.
0
50.
0
60.
0
70.
0
80.
0
A i Tem p. 0.
r
( 1℃)
R el onshi betw een S urf Tem p and A i Tem p
ati
p
ace
r
90.
0
100.
0
110.
0
120.
0
76. Microwave Radiometer
- Passive Microwave Sensor
Measuring radiated microwave from an object
AMSR-E Instrument
Description
http://www.ghcc.msfc.nasa.gov/AMSR/html/amsr_products.html
The PM-1 AMSR is a twelve
channel, six frequency total power
passive microwave radiometer
system. It measures brightness
temperatures at 6.925, 10.65, 18.7,
23.8, 36.5, and 89.0 GHz
83. AMSR-E Level 2 EOS Standard Data Products
PARAMETER
ACCURACY
SPATIAL
RESOLUTION
Brightness Temperature
0.2 - 0.7 K
6 - 76 km
Ocean Wind Speed
1.5 m/s
12 km
Water Vapor
Over Ocean
0.2 g/cm2
23 km
3 mg/cm2
23 km
0.5 K
76 km
Surface Soil Moisture
0.06 g/cm3
where vegetation is less
than 1.5 kg/m2
25 km
(Equal Area Earth
Grid)
Global Rainfall
Ocean: 1 mm/hr or 20%
(whichever is greater)
10 km
Global Rain Type
(Convection fraction)
Land: 2 mm/hr or 40%
(whichever is greater)
N/A
10 km
Cloud Liquid Water
Over Ocean
Sea Surface Temperature
88. Landsat 1 to 8
Satellite
Launch Date
Landsat 1
Period of Operation
23 July 1972
Landsat 6
Decommissioned 6
January 1978
Decommissioned 25
22 January 1975
February 1982
Decommissioned 31
5 March 1978
March 1983
Decommissioned
16 July 1982
June 2001
Thematic Mapper
stopped acquiring
1 March 1984
data 18 November
2011
October 1993
Failed on Launch
Landsat 7
15 April 1999
Operating in SLC-Off
Mode after May 2003
Landsat 8
February 2013
Due to be launched
February 2013
Landsat 2
Landsat 3
Landsat 4
Landsat 5
(1972 – present)
http://www.ga.gov.au/ausgeonews/ausgeonews2012
09/landsat.jsp
92. SPOT-5 sample image of Naples (Italy) in 2002 (image credit: CNES)
the spatial resolution of the imagery to < 3 m in the panchromatic band and to 10 m in the multispectral mode
https://directory.eoportal.org/web/eoportal/satellite-missions/s/spot-5
93. MOS-1
Main Characteristics of the MOS-1
------------------------------------------Scape
: Box type with expanding type
solar cell paddle (one wing)
Bus unit
1.26mx2.4mx1.48m
Solar cell paddle, total length 5.28mx2m
Weight
: Approx. 740kg
Attitude control : Three axes control
Design life
: 2 years
------------------------------------------Launch vehicle : H-I
Launch site : Tanegashima Space Center,
Kagoshima
Launch date : February 7, 1990
------------------------------------------Orbit Type : Sun synchronous subrecurrent orbit
Altitude : Approx. 909km
Inclination : Approx. 99deg.
Period
: Approx. 103min.
99. ADEOS
Sensors in ADEOS
1. OCTS - Ocean Color and Temperature
Scanner
2. AVNIR - Advanced Visible and Near
Inrared Radiometer
3. NSCAT - NASA Scatterometer
4. TOMS - Total Ozone Mapping
Spectrometer
5. IMG - Interferometric Monitor for
Greenhouse Gases
6. POLDER - Polarization and
Directionality of the Earthe's Reflectance
7. ILAS - Improved Limb Atmospheric
Scatterometer
112. ASTER G-DEM
International joint project between METI and NASA
Earth observing sensor developed by Japan (METI) flying on Terra
Launched in December 1999, in stable operation for more than 7 years
ASTER provides:
Flight direction
Backward
satellite Terra
Nadir
1) Surface condition
The earth surface is observed in visible to
thermal infrared (invisible to human eyes)
spectral regions to obtain detailed
information on the condition and
distribution of the surface
(vegetation, geology, etc.).
2) Surface temperature
The distribution of surface temperature is
observed by the thermal infrared sensor to
study the urban heat island effect and
other phenomenon in detail.
3) DEM
DEM is derived from a stereo-pair of
images over a single area acquired in nadir
and backward viewing angles.
113. Features of ASTER G-DEM
Joint project between METI and NASA
Generation of global land DEM based on the ASTER coverage
Enhanced accuracy due to the use of multiple ASTER data over one region
User friendly with the capability for selective cropping
ASTER scene
(60km x 60km)
Generation of seamless
DEM using all ASTER data
ever acquired over the
target area
Automated processing
A seamless wide-coverage
Red-colored area: ASTER coverage (1.1 million scenes)
Deeper red indicates more frequent observations,
thus providing higher accuracy.
ASTER G-DEM
DEM
applied to all land area
Easy to use, allowing for
selective cropping
114. Comparison with other DEMs
ASTER G-DEM
SRTM3
GTOPO30
Shuttle Radar Topography
Mission Data at 3 Arc-Seconds
Global 30 Arc-Second
Elevation Data Set
Data source
ASTER
Space shuttle radar
From organizations around the
world that have DEM data
Generation and
distribution
METI of Japan / NASA
NASA/NGA/USGS
USGS
Release year
2009 ~ (planned)
2003 ~
1996 ~
Data acquisition period
2000 ~ ongoing
11 days (in 2000)
DEM resolution
30m
90m
1000m
DEM accuracy
(stdev.)
±7m
±10m
±30m
DEM coverage
83 degrees north
~83 degrees south
60 degrees north
~56 degrees south
Global
Area of missing data
Areas with no ASTER data due
to constant cloud cover
Topographically steep area
None
(due to radar characteristics)
NGA:National Geospatial-intelligence Agency
USGS:United States Geological Survey
The ASTER G-DEM is the only sophisticated global coverage DEM, which will be widely used as
the global standard.