Image Processing Reference
In-Depth Information
•
ERDAS Imagine (Leica Systems).
http://www.gis.leica-geosystems.com/
products/imagine/
IDRISI Kilimanjaro (Clark Labs).
•
http://www.clarklabs.org/
eCognition 3.0 (Definiens Imaging GmbH). Free evaluation available.
•
http://
www.definiens-imaging.com/
Study Questions
•
How many practical urban applications would benefit from the classification of
satellite sensor data?
Compare and contrast land cover and land use with respect to various urban
•
applications.
Discuss the dilemma of, on the one hand, increasing spatial resolution and, on
•
the other, decreasing classification accuracy in the interpretation of urban cate-
gories from satellite sensor data.
Do you think spectral-based classification or spatial-based pattern recognition
•
are more intuitive in how urban areas are structured and how processes are
interpreted?
How processes are interpreted?
•
LAB EXERCISE
You will need to install ENVI Guided Evaluation software from the following RSI
link http://www.rsinc.com/envi/ and download a fully functional version which
runs in restricted sessions of 7 min. You will also need to save the image
belfast_
subset
from the Topic's homepage at springer.com (http://www.springer.com/sgw/
cda/frontpage/0,11855,4-40106-22-107950926-0,00.html) to your local hard disc.
This is a small subset of a pan-sharpened 4 m multispectral IKONOS image of the
city of Belfast in Northern Ireland (alternatively you can follow the same steps
using an image subset of your own city).
•
Data import
: Start ENVI Evaluation version. From
File
choose
Open Image File
and select
belfast_subset
. Another window will open,
Available Bands List
.
Click on
RGB Color
and make sure that
Layer 4
is loaded in
Red
,
Layer 3
in
Green
and
Layer 2
in
Blue
. Click on
Load RGB
.
•
Display
: Two other windows of differing resolutions will pop open. You may
synchronize these windows to zoom in and out of the image. The subset matches
the configuration of built structures as used in Figs.
8.4
and
8.5
. Notice the red
areas correspond to vegetation, the linear arrangement of the roads is also dis-
cernible but the buildings vary in composition and are difficult to delineate.