CS 128/ES 228 - Lecture 10a 1 Raster Data Sources: Paper maps & Aerial photographs
Jan 20, 2016
CS 128/ES 228 - Lecture 10a 1
Raster Data Sources:Paper maps &
Aerial photographs
CS 128/ES 228 - Lecture 10a 2
Georgian’s First Law of GIS
Try to use somebody else’s
data before you even think of
generating your own.
CS 128/ES 228 - Lecture 10a 3
Data sources: overview
Raster sources: Paper maps Aerial photographs Satellite images
Vector sources: Digitized maps Surveying Global positioning system
CS 128/ES 228 - Lecture 10a 4
Paper maps
CS 128/ES 228 - Lecture 10a 5
Using paper maps in a GIS
1. Scan to image file (usually JPEG or GIF)
2. Georeference the image to the GIS coordinate system
3. (If desired) digitize the features in the image to generate 1+ vector layers
CS 128/ES 228 - Lecture 10a 6
Georeferencing raster images Spatial coordinates may be absent or purely
map coordinates (i.e. inches from one corner)
Control points: point features visible on both the image and the map
Linear or nonlinear transformations
“Rubber sheeting”
CS 128/ES 228 - Lecture 10a 7
Affine transformations
Translation
Rotation
Scaling
Skew
CS 128/ES 228 - Lecture 10a 8
Georeferencing in ArcMap - 1
CS 128/ES 228 - Lecture 10a 9
Georeferencing in ArcMap - 2
CS 128/ES 228 - Lecture 10a 10
Georeferencing in ArcMap - 3
CS 128/ES 228 - Lecture 10a 11
Georeferencing in ArcMap - 4
CS 128/ES 228 - Lecture 10a 12
Georeferencing in ArcMap - 5
CS 128/ES 228 - Lecture 10a 13
Georeferencing in ArcMap – 5b
CS 128/ES 228 - Lecture 10a 14
Georeferencing in ArcMap - 6
CS 128/ES 228 - Lecture 10a 15
Georeferencing in ArcMap - 7
CS 128/ES 228 - Lecture 10a 16
Georeferencing in ArcMap - 8
CS 128/ES 228 - Lecture 10a 17
Georeferencing in ArcMap - 9
CS 128/ES 228 - Lecture 10a 18
Digitizing raster layersDigitizing table high resolution (0.001”) either point or stream
mode paper shrinkage/
expansion data in “table
coordinates” – need to convert to map coordinates
CS 128/ES 228 - Lecture 10a 19
“Heads up” digitizingTracing on computer
monitor: many scanned (raster)
file formats supported poorer resolution, but
uses less specialized equipment
best for adding small # features or updating a file
uses coordinate system of image or base map
CS 128/ES 228 - Lecture 10a 20
Aerial photographs
HUGE amount of detail
VAST number of photographs are available, often for free
Digitizing and photo-interpretation can produce vector layers and attribute data
CS 128/ES 228 - Lecture 10a 21
Photogrammetry Originally, the science
(or art?) of interpreting aerial photographs
Stress on quantitative measurements
Now includes analysis of digital images from many sources
Image from Avery. Interpretation of Aerial Photographs.
CS 128/ES 228 - Lecture 10a 22
ScaleDetermine from: Plane altitude
RF = lens focal length altitude of plane
Known ground features
Top image from Avery. Interpretation of Aerial Photographs.Bottom images from Ben Meadows catalog (L), Olean NW DOQQ ®
CS 128/ES 228 - Lecture 10a 23
Perspective Vertical:
- orthogonal perspective- planimetric map data
Oblique: - high oblique
(includes horizon) - low oblique (no horizon)
Image from Avery. Interpretation of Aerial Photographs.
CS 128/ES 228 - Lecture 10a 24
Planimetric view
Perfectly vertical (orthogonal) perspective
All features in correct horizontal positions
Impossible unless at infinite height
CS 128/ES 228 - Lecture 10a 25
The principle point
Point directly under camera lens (‘nadir’)
Elevated objects lean away from PP
Depressed objects lean toward PP
Causes horizontal image displacement
Images from Avery. Interpretation of Aerial Photographs.
CS 128/ES 228 - Lecture 10a 26
Vertical relief -> displacement Transmission line is
straight - why does the line appear straight in one photo and jagged in the second?
In left photo, line is ~ on nadir; in right photo, the line is far from nadir
Image from Avery. Interpretation of Aerial Photographs.
CS 128/ES 228 - Lecture 10a 27
Image displacement:
Source of error in horizontal locations, but
Permits estimation of feature elevations
stereoscopic parallax
Image from Avery. Interpretation of Aerial Photographs.
CS 128/ES 228 - Lecture 10a 28
Stereoscopic photo pairs
Image from Avery. Interpretation of Aerial Photographs.
CS 128/ES 228 - Lecture 10a 29
Stereoscopes
need pair of overlapping photos
different principle points results in parallax
used to create topographic contours
CS 128/ES 228 - Lecture 10a 30
Rectification of aerial photographsRectification: process of geometric correction
that turns an aerial photograph into a planimetric (map-like) image
Problems: Earth curvature lens distortion camera tilt terrain relief
CS 128/ES 228 - Lecture 10a 31
Rectification process
1. Scan aerial photograph at high resolution
2. Locate ground control points on scanned image: ≥3 for affine transformation ≥5 for rubbersheeting
3. Combine with digital elevation model (DEM) to correct relief displacement
4. Rectify to a ground coordinate system
CS 128/ES 228 - Lecture 10a 32
Relief distortion
Objects at different distances form the lens will be distorted
CS 128/ES 228 - Lecture 10a 33
Urban areas: building tilt
In urban areas, tall buildings seem to lean toward the principal point of the photograph
Corrected by building a digital terrain model (DTM) of each building
Permits virtual reality “flyovers”
Thorpe, A. Digital orthophotography in New York City. www.sanborn.com/Pdfs/Article_DOI_Thorpe.pdf
CS 128/ES 228 - Lecture 10a 34
Result: digital orthophotograph
USGS: DOQQs
NYS GIS Clearinghouse
Or, new aerial photos & image rectification ($$$)
Wind Cave N P Vegetation Survey? Sure (tax $$)
No!!!CS 128/ES 228 Course Project?