Land Cover Mapping Tutorial

Training Manual
Land cover mapping

using satellite data
Draft
Contents
1. Introduction ............................................................................... 1
1.1
Legend development and classification scheme .............................................. 3
1.2
Available satellite images and their spatial and spectral resolutions .................. 4
2. Fundamental concepts in satellite image classification ............... 6

2.1
Image ratios for information extraction .......................................................... 7
3. Methodology of image analysis for land cover mapping ........... 8

3.1
Image rectification ...................................................................................... 8
3.2
Image classification ..................................................................................... 9
3.3
Object-based image analysis (OBIA) ........................................................... 10
3.4
Accuracy assessment ................................................................................. 12
3.5
Land cover change analysis ....................................................................... 13
4. Hands on Exercise Using eCognition Developer ....................... 18

4.1
Getting started .......................................................................................... 18
Create a New Project.......................................................................................... 19

Subset Selection.................................................................................................. 21
Insert Thematic Layer ........................................................................................... 22
4.2
Image Objects by Segmentation ................................................................. 24
4.3
Classification of Land Cover Using Landsat ETM+ Image ............................... 25
View Settings Toolbar.......................................................................................... 27

Insert Rule for Object Creation ............................................................................. 28
Create Relational Feature..................................................................................... 31
Land & Water Mask (LWM) ................................................................................. 32
Insert the Class/Class Hierarchy ........................................................................... 33
Insert a Classification Parent Process ..................................................................... 33
4.4
Manual Editing ......................................................................................... 60
4.5
Annex I: Information on Landsat and Indices ................................................ 66
5. References and Useful links ...................................................... 68
Land cover mapping using satellite data
ii
1. Introduction
The Hindu Kush-Himalaya (HKH) is the youngest and one of the most fragile mountain
ecosystems in the world. The HKH region encompasses about 3500 kilometers east to west
and includes eight regional member countries (RMCs) namely- Afghanistan, Bangladesh,
Bhutan, China, India, Myanmar, Nepal and Pakistan. It is the youngest, highest and one
of the most fragile mountain systems of the world. It is estimated that more than 1.3 billion
people are directly dependent upon the ecosystem services provided by these mountains.
Most mountain communities are dependent upon subsistence agriculture and natural
resources. More recently, climate change has put the Himalayan region in centre of the
international attention - as one of the most vulnerable ecosystems in the world severely
impacting its social and environmental security of the region. The vulnerability of the HKH
region, to the effects of socio-economic development in the region together with
degradation caused by the improper use of natural resources, becomes a critical element
which needs to be assessed in the planning and management of sustainable mountain
development in the region. The best indicator to monitor these vulnerabilities is land cover
and use, and its dynamics over time.
Distribution and changes in land cover affect the ecosystem services (e.g., provision of
food and fiber, sustaining biodiversity, providing recreational places etc.) of an area,
induce climate changes by modifying water and energy exchanges with the atmosphere,
and distort greenhouse gas balances. Thus land cover information is a key input to a wide
range of interventions on issues of national to global interest, including land degradation,
climate change, food security, poverty and environmental sustainability. In addition many
of the climate variables that are difficult to measure at large scale can be partly inferred
by interpreting the vegetation and land-surface types. Thus, land cover can serve as proxy
to other important climate variables.
Land cover refers to the physical and biological cover over the surface of land, including
water, vegetation, bare soil, and/or artificial structures. Land use denotes how humans use
the biophysical or ecological properties of land. Land use is characterized by the
arrangements, activities and inputs people undertake in a certain land cover type to
produce, change or maintain it. Definition of land use in this way establishes a direct link
between land cover and the actions of people in their environment. Information on land
use and land cover is required in many aspects of sustainable management of land
resources and policy development, as a prerequisite for monitoring and modelling land
use and environmental change, and as a basis for land-use statistics at all levels (Jansen &
Di Gregorio 2004).
Land cover-land use analysis provides knowledge about landscape patterns and their
changes which over time gives very important insights into the ongoing natural and human
processes in the ecosystem. Human activities are a major factor contributing to global
change, and they are overriding natural changes to ecosystems brought on by climate
variations (Figure 1). On the local level land use changes can fundamentally alter the
availability of natural inputs to ecosystems (energy, water, and nutrients), generate exotic
1
species invasions, and accelerate natural processes of ecosystem change (Ojima et al.,
1994).
Figure1: Processes and impact of land use changes (Ojima et al. 1994)
The demand of information on land cover, land use and their changes is increasing at the
global, regional and national levels to support policy decisions and regulate management
processes.
In contrast to survey data, aerial and satellite imagery can be used to monitor the spatial
extent of changes in land cover (i.e., conversion) or land conditions. Satellite imagery
offers contiguous spatial coverage, facilitates better repetition and replaces costly and
slow data collection of different land cover and provides statistical information of the area
/object. Remote sensing technologies have made substantial contribution in deriving land
cover information and correlating to land-use statistics (Jansen &Di Gregorio 2004).
Availability of satellite images with different spatial and temporal resolutions have made it
possible to map land cover at different scales and carry out analyses of the changes over
last three decades. Specifically, remote sensing change detection analyses can be used to
identify areas of rapid change to target management efforts (Rogan et al 2002; Coppin et
al 2004; Kennedy et al 2009). Repeated satellite images and/or aerial photographs are
useful for both visual assessment of natural resources dynamics occurring at a particular
time and space as well as quantitative evaluation of land cover changes (Tekle and
Hedlund 2000). A number of deforestation and degradation studies have been conducted
in tropical forests using coarse and high-resolution remote sensing data (Qamer et al
2013, Panta et al 2008; Gautam et al 2004; Jha et al 2000; Skole and Tucker 1993).
The temporal evaluation of forest changes based on satellite imagery is becoming a
valuable set of technique for assessing the degree of threat to ecosystem (Giriraj et al
2
2008; Reddy et al 2008). GIS on the other hand provides environment to analyze digital
data useful for change detection, database development, and modeling of its future
change and data dissemination for effective management planning. In context of Reducing
Emissions from Deforestation and Degradation (REDD), Remote-sensing methods are
considered to be appropriate for most developing countries to assess historical and future
deforestation rates, i.e., forest area change (GOFC-GOLD 2009).
Study of land cover dynamics involves developing a multi temporal land cover database
using satellite images of different dates. In this context, a uniform/comparable land cover
legend is essential for meaningful comparisons between different time periods. The land
cover mapping have to be useful for applications at different scales and therefore it is
important to design a system which follows a uniform approach and allows aggregation at
different levels of detail. Harmonization of the classification system is therefore an
important step; it will not only facilitate the generation of land cover maps that are
compatible with each other but also can be used consistently for change studies. Land
cover mapping requires significant resources and due to the gaps in harmonized legends
and methodologies, investments in past initiatives could not be properly used for change
studies which are crucial for decision support for key applications. In order to make the
current efforts sustainable, the capacity building of the partners in the regional member
countries is an important step to utilize common approach and methodology to develop
regional land cover database periodically. Such a database is crucial for assessing key
drivers for land cover and land use change and decision-making.
1.1
Legend development and classification scheme
The definition of land cover is fundamental, because in many existing classifications and
legends it is confused with land use. A classification describes the systematic framework
with the names of the classes and the criteria used to distinguish them, and the relation
between classes. Classification thus necessarily involves definition of class boundaries that
should be clear, precise, possibly quantitative, and based upon objective criteria.
Understanding of land use and cover is essential to understand landscape patterns and
their change in past and for that harmonizes and standardizes classification is very
important (Birendra et al 2010). In order to address these differences, a number of
organizations and institutions are working to create general classification systems and
legends for global consistency, such as terrestrial ecoregion. To address this urgent need,
the Food and Agriculture Organization (FAO) has developed a system for land cover
classification (Di Gregorio 2005). The International Centre for Integrated Mountain
Development initiated research to harmonize land cover classification at the regional scale
and address the immediate needs of the Hindu Kush Himalayan Region. Harmonization of
the classification system is facilitating the generation of land cover maps that can be used
consistently for studies of change. The table 1 below shows example harmonized land
cover legend developed using a Land Cover Classification System (LCCS).
3
Table 1: legend of land cover for different levels
Level-1
Cultivated
Managed
Agriculture
Level-2
Valley
cultivation
Level terrace
Needleleaved
forest
Forest
Primary
vegetated
Area
Terrestri
al
Broadleaved
forest
Mixed Forest
Natural
and Semi
Natural
Vegetation
Needleleaved
Shrub
Shrub land
Broadleaved
shrub
Mixed Shrub
Primary
non
vegetated
Area
Artificial
surface
Bare area
Aquatic
/Regular
ly
Flooded
Natural
Waterbodi
es, snow
and ice
Mixed crops
Paddy
Current fallow
Needleleaved open
forest
Needleleaved closed
forest
Broadleaved closed
forest
Broadleaved open
forest
Mixed open forest
Mixed closed forest
Needleleaved open
Shrub
Needleleaved closed
Shrub
Broadleaved closed
shrub
Broadleaved open
shrub
Mixed closed shrub
Mixed open shrub
Settlement
Road
Bare rock
Bare Soil
River
River bed
Needleleaved
grass
Broadleaved
grass
Settlement
Road
Bare rock
Bare Soil
River
River bed
Settlement
Road
Bare rock
Bare Soil
River
River bed
Snow/ice
Snow/ice
Snow/ice
Grass land
Terrestri
al
Level-3
Needleleaved grass
Broadleaved grass
1.2 Available satellite images and their spatial and spectral

resolutions
Land cover maps derived from satellite images play a key role in global, regional, national
and subnational land cover assessments. Land cover map is available in a range of data
formats and spatial resolutions to suit different user requirements. According to the study
area size and details of mapping selection suitable satellite image are important. Presently
satellite imagery became widely available when affordable. The relevance of different
4
satellites and information which can be generated to suit different scales of land cover
assessment are given in Table- 2.
Table 2: Commonly used satellite imagery with resolution and scale
Sl
No
Launch
Satellite/
Sensor
Band/Resolution
Scale
Leve
l*
Jun 2014
WorldView2
Very high-resolution with 8 Band

(Pan and Multi)*
- Panchromatic 31 cm
- Multispectral 1.24 m
- Short-wave infrared 3.7 m
5000
Oct 2009
WorldView2
5000
Oct 2001
Quickbird
Very high-resolution with 8 Band

(Pan and Multi)*
-Panchromatic 46 cm
-Multispectral 1.85 m (red, blue,
green, near-IR, red edge,
coastal, yellow, near-IR2)
High-resolution with 5 Band
(Pan and Multi)*
- Panchromatic 61 cm
- Multispectral 2.44 m
10000
Sep 2009
Cartosat-2
High-resolution
-Panchromatic 1 m
10000
Jan 2006
ALOS(PRISM,
High-resolution
-Panchromatic 2.5m
-Multispectral 10m
50000
Oct 2003
IRS LISS IV
MX
High medium resolution

-Multi (Green, Red and NIR)
5.8 m
50000
Apr 2009
Landsat7
ETM+
Medium-resolution with 8 Band

(Pan and Multi)*
-Panchromatic 150m
-Multispectral 30m (TR 60m)
100000
Apr 2009
Landsat5 TM
Medium-resolution with 7 Band

-Multispectral 30m (TR 120m)
100000
AVNIR,
PALSAR)
* Land cover map legends at different scales
Quicklook
2. Fundamental concepts in satellite image classification

The basic principle behind these assessments based on passive remote sensing is
exploitation of electromagnetic spectrum. Energy from the sun comes in many different
wavelengths. Various features on earths surface respond differently to different
wavelengths. In simple feature has its own "signature".
Different surface types such as water, bare ground or vegetation reflect radiation
differently in various channels. The radiation reflected as a function of the wavelength is
called the spectral signature of the surface.
A. Graphs of spectral signatures of water, soil and vegetation. Vegetation has a
remarkably high reflection in the near infrared channel 4 and a low reflection in the visible
red channel 3. This makes it possible to distinguish vegetation areas from bare ground.
The difference of reflection in channels 3 and 4 is great for vegetation areas and
insignificant for bare ground.
B. The spectral signatures are processed as digital values in the satellite scanner. Here is a
hypothetical example of how the LANDSAT satellite might record water, green vegetation
and bare ground.
The reflection from bare ground increases slightly from the visible to the infrared range of
the spectrum. There are great differences between different types of soil, dry and humid
land. Different mineral compositions of the surface are also reflected in the spectral
signature.
Generally, water only reflects in the visible light range. As water has almost no reflection
in the near infrared range it is very distinct from other surfaces. Thus water surfaces will be
clearly delimited as dark areas (low pixel values) in images recorded in the near infrared
range.
The spectral signature for green plants is very characteristic. The chlorophyll in a growing
plant absorbs visible and especially red light to be used in photosynthesis, whereas near
infrared light is reflected very effectively as it is of no use to the plant, see illustration. In
this way the plants avoid unnecessary heating and loss of juice through evaporation.
Therefore the reflection from vegetation in the near infrared and in the visual ranges of the
spectrum varies considerably. The degree of difference reveals how large a part of the
area is covered with growing green leaves (leaf area index).
Different surface types such as water, bare ground or vegetation reflect radiation
differently in various channels. The radiation reflected as a function of the wavelength is
called the spectral signature of the surface.
2.1
Image ratios for information extraction
Image ration is very important to during the image classification. Image rationing is a
synthetic image layer created from the existing bands of a multispectral image. This
new layer often provides unique and valuable information not found in any of the other
individual bands. Image index is a calculated results or generated product from satellite
band/channels. It is help to identify different land cover from mathematical definition.
Normalized difference vegetation index (NDVI)

One of the commonly used indices is NDVI. NDVI is related to vegetation is that healthy
vegetation reflects very well in the near infrared part of the spectrum. NIR stands for Near
Infrared.
NDVI = (NIR - red) / (NIR + red)
Figure 2: Satellite image and NDVI of

Sagarmatha National Park
Figure 1: Normalized difference vegetation index of ETM+ image
NDVI index values can range fro
NDVI index values can range from -1.0 to 1.0, but vegetation values typically range
between 0.1 and 0.7. Free standing water (oceans, seas, lakes and rivers) which have a
rather low reflectance in both spectral bands and thus result in very low positive or even
slightly negative NDVI values. Soils which generally exhibit a near-infrared spectral
reflectance somewhat larger than the red, and thus tend to also generate rather small
positive NDVI values (say 0.1 to 0.2).
Normalized Difference Snow Index (NDSI)

Difference Snow/Ice Index calculations are related to reflections different bands. Snow
and ice have very high reflectance values in visible spectral bands (blue, green and red),
7
but very low reflectance in mid-infrared band. The value is then normalized to the range 1<=NDVI<=1 to partially account for differences in illumination and surface slope. All the
snow will carry positive value.
NDSI = (green IR) / (green + IR)
3. Methodology of image analysis for land cover mapping
The generic steps followed in land cover change assessment using satellite is given below
Image classification scheme

Data acquisition
Image rectification and
enhancement
Field training information
Image segmentation
Generate image indices
Assign rules
Land cover map
Accuracy assessment and validation
Finalization of land cover
Figure 3: Flow diagram of land cover mapping
3.1
Image rectification
Pixels on raw satellite remote sensing images only have row, column coordinates; that is,
they do not have geographic coordinates such as latitude-longitude or state plane
coordinate system or raw imagery has no reference to the ground. Remote sensing images
display varying degrees of geometric and location distortion. The rectification of an image
transforms it to display a plane object in the image as if the picture has been taken directly
8
from the satellite. Rectification is the process of giving an image a real World coordinate
system. After rectification the scale of the image is uniform, and it's the real scale,
distances on the object can be measured directly on the picture.
3.2
Image classification
Image analysis is the extraction of meaningful information from images; mainly from
digital images by means of digital image processing techniques. The process of sorting
pixels into a number of data categories based on their data file values and reducing
images to information classes. Image classification refers to the task of extracting
information classes from a multiband raster image. The resulting raster from image
classification can be used to create thematic maps. Image classification in the spectral
domain is known as pattern recognition in which the decision rules are based solely on the
spectral values of the remote sensing data. In spatial pattern recognition, the decision rules
are based on the geometric shape, size, texture, and patterns of pixels or objects derived
from them over a prescribed neighborhood. In order to take advantage of and make good
use of remote sensing data to developing land cover, extract meaningful information from
the imagery is very important. During the image classification following interpretation
factors need to consider:
Tone: Tone refers to the relative brightness or color of objects in an image. Generally,
tone is the fundamental element for distinguishing between different targets or features.
Variations in tone also allow the elements of shape, texture, and pattern of objects to be
distinguished.
Shape: Shape refers to the general form, structure, or outline of individual objects. Shape
can be a very distinctive clue for interpretation. Straight edge shapes typically represent
urban or agricultural field targets, while natural features, such as forest edges, are
generally more irregular in shape,
Size: Size of objects in an image is a function of scale. It is important to assess the size of
a target relative to other objects in a scene to aid in the interpretation of that target.
Pattern: Pattern refers to the spatial arrangement of visibly discernible objects. Typically an
orderly repetition of similar tones and textures will produce a distinctive and ultimately
recognizable pattern. Orchards with evenly spaced trees, and urban streets with regularly
spaced houses are good examples of pattern.
Texture: Regular arrangement of ground objects. Examples are urban area and rural
agriculture area arrangement on satellite imagery.
Shadow : Shadow is also helpful in interpretation as it may provide an idea of the profile
and relative height of a target or targets which may make identification easier. Example is
on the very high resolution trees will have shadow and shrub will have without shadow.
Association: Association takes into account the relationship between other

recognizable objects or features in proximity to the target of interest. The
identification of features that one would expect to associate with other features
may provide information to facilitate identification.
Depending on the interaction between the analyst and the computer during classification,
there are different types of classification procedures:
3.3
Unsupervised
Supervised
Knowledgebase
Object base
Others
Object-based image analysis (OBIA)
Classification is producing meaningful material distribution maps via identification of

individual pixels or groups of pixels with similar spectral responses to incoming radiation.
These pixels or groups represent different materials or classes. Object-Based Image
Analysis (OBIA) also called Geographic Object-Based Image Analysis (GEOBIA)". OBIA
is a sub-discipline of geoinformation science devoted to partitioning remote sensing
imagery into meaningful image-objects, and assessing their characteristics through spatial,
spectral and temporal scale". The fundamental step of any object based image analysis is
a segmentation of a scenerepresenting an imageinto image objects. Thus, initial
segmentation is the subdivision of an image into separated regions represented by basic
unclassified image objects called Image Object Primitives.
Typically; an Object Based Image Analysis starts with a crucial step of Image
Segmentation in which meaningful image object are created then these image objects are
classified in the later step of classification. In the field of forestry; object-based techniques
such as per-field or per-parcel classification (Aplin et al., 1999); and point-based
segmentation algorithm (Heyman et al. (2003),) have been evolved to delineate imageobjects or zones of contiguous pixels belonging to same class. Object-based image
analysis provides significant information about forest ecosystem (Mallinis et al., 2008).
Initially, image objects are constructed by image segmentation, which is the subdivision of
an image into separate multi-pixel regions. (Benz et al. 2004). The most frequently used
segmentation algorithm Multiresolution Segmentation has been adopted to generate
image object (Kozak et al., 2008; Mallinis et al., 2008; Mathieu et al. 2007; Mather,
2004; Flanders et al., 2003; Manakos et al. 2000; Hay et al., 2005; Santos et al.,
2006). Multiresolution segmentation algorithm is based on a pair wise region merging
technique in which pixels (one pixel object) or existing image object are successively
10
merged (Zhou et al., 2008). Definiens (eCognition) provides segmentation on several

scales that is very important for a meaningful image analysis. (Benz et al. 2004).
Segmentation on several scales with different scale parameters can be carried out leading
to the formation of a hierarchical network of objects. This procedure is constrained so that
spatial shape of objects in one level fits hierarchically into objects of another level
enabling consideration of sub-objects and super-objects and their mutual relationships in
the classification step (Mallinis et al., 2008).
Segmentation
Vegetated
Build
classification
Hierarchy
Farmland
Non-Vegetated
Identification
based on indices
Natural
Bare
vegetat
ion
Area
Water/
Snow Area
Classification
Vegetated
Classification-based
Segmentation
Farmland
NonVegetated
Natural
vegetat
ion
Identification
based on
scale, shape
Bare
Area
Classification
Accuracy
Assessment
Figure 4: Flow diagram of Hierarchal Object based Image classification scheme
11
Water/
Snow
Area
Classification Assumptions
Similar features will have similar spectral responses.

The spectral response of a feature is unique with respect to all other features of
interest.
If we quantify the spectral response of a known feature, we can use this
information to find all occurrences of that feature.
Figure 5: Class Hierarchy and respective classified objects
3.4
Accuracy assessment
Accuracy assessments determine the quality of the information derived from remotely
sensed data. The product of image classification is land cover maps. Their accuracy needs
to be assessed so that the ultimate user is made aware of the potential problems
associated with their use. Accuracy assessment is a quality assurance step in which
classification results are compared with what is there on ground at the time of imaging or
something that can be regarded as its acceptable substitute, commonly known as the
ground reference. Evaluation of the accuracy of a classification may be undertaken for
each of the categories identified and its confusion with other covers, as well as for all the
12
categories. The outcome of accuracy assessment is usually presented in a table that

reveals accuracy for each cover category and for all categories as a whole.
A sample accuracy assessment report
3.5
Land cover change analysis
Reliable information on land cover (LC) and land-cover change is required for a wide
range of environment-related topics and activities, including forest management, which has
changed dramatically in the last years. Determining the effects of land-use and land-cover
change on the Earth system depends on an understanding of past land-use practices,
current land-use and land-cover patterns, and projections of future land use and cover, as
affected by human institutions, population size and distribution, economic development,
technology, and other factors. The combination of climate and land-use change may have
profound effects on the habitability of Earth in more significant ways than either acting
alone. While land-use change is often a driver of environmental and climatic changes, a
changing climate can in turn affect land use and land cover. Climate variability alters landuse practices differently in different parts of the world, highlighting differences in regional
and national vulnerability and resilience.
Land-cover changes can be driven by anthropogenic and natural alterations. Alterations in
LC affect the regional or global energy balance, hydrologic cycle, biogeochemical cycles,
and climate. Land-cover change can be driven by:
13
climate-related change (both long and short term);

burning whether deliberate or wildfire;
cutting and clearing of forests (deforestation);
grazing activities (intensification of rangeland use);
agricultural encroachment (farming activities)

abstraction for fuelwood;
urban expansion (urbanization);
It is possible to monitor human and natural changes in the landscape, such as

deforestation, change in agricultural land use pattern, desertification, flooding, soil
erosion, plant community change and sub-urbanization or urbanization, using remotely
sensed products where the detection of changes can be highlighted using cyclical satellite
passages. The ability to detect change is, in fact, one of the benefits of remote sensing (RS)
because the use of multitemporal data sets permits discrimination of areas of change
between imaging dates.
Change detection analysis with RS depends upon the ability to isolate one type of
change from among numerous changes taking place at different temporal and spatial
scales (Lillesand, Kiefer and Chipman, 2004). A fundamental assumption in digital change
detection is that a difference exists in the spectral response from two dates if the
biophysical material within the Instantaneous Field of View (IFOV) has changed between
dates (Jensen, 1996).
Change detection involves the use of multitemporal or multi-type data sets, or a
combination, to discriminate areas of Land Cover between dates of imaging. It requires a
very accurate choice of products to be used for the specific scope, and consequently the
images chosen must undergo accurate co-registration. The land cover change
interpretation process must result in a systematic description of the changes present in the
study area. Change characterization can mean not only identifying intensification or
reduction of LC and other land use, but also evaluating the type and the main peculiarities
of changes so as to be able to understand the induced consequences on the environment.
Characterizing change has certain implications, considered below.
Analysis of present features: The identification of the present LC features can be
considered the first task in the process.
Detecting areas of change: The identification of areas of land-cover change in
relation to a certain previous period is the second task of the research. The area of change
must be delimited in order to define the exact extension and the correct location of
changes in the study area.
Magnitude of change: Together with the physical extent, the intensity of change must
be defined. Magnitude is quite difficult to quantify, but, simplistically, two main categories
can be used for identification purposes, namely change and modification.
Three main factors control change detection accuracy.
Spatial properties: Image resolution controls the minimum size of detectable

changes determined by the IFOV. Also, the quality of the geometric registration
controls the minimum size of detectable changes by adding an error term to
resolution. It is strongly affected by pixel size.
14
Radiometric and spectral properties (bandwidth and radiometric

resolution): Contrast between cover types is dependent on bandwidth and location
of the bands. These factors delimit the types of land-cover changes that can be
detected. Radiometric resolution controls the size of radiance differences that are
detectable.
Temporal properties: The time interval between images and frequency of

imaging must take into account not only orbiting frequency but also cloud and
haze cover frequency. Moreover, the re-visit interval affects how small a change
can be detected (i.e. the size of changes it is possible to detect).
15
16
Hands on Exercise
17
4.
Hands on Exercise Using eCognition Developer
4.1
Getting started
Go the Windows Start menu and Click Start > All Programs> eCognition Developer
8.0> eCognition Developer
Upon launching Definiens eCognition Developer 8, the following dialog appears:
18
Figure: eCognition 8 launch screen
Figure 6: The default display eCognition 8
Create a New Project

To Create New Project do the following:
Choose File > New Project on the main menu bar.
19
Navigate the folder C:\GISRS_Trn\Definiens

Select Image.img > Open (Here is image file Landsat ETM+, R136/P44)
Then select from the appropriate file in the files type.
To open some specific file formats or
structures, you have to proceed as follows:
First select the correct driver in the Files
of type drop-down list box
Double-Click on Image Layer Alias
Rename the all layers name
Click> OK
Click> Insert > Select DEM.img and
Slope> Open
20
Double-Click on Layer Alias Rename the all the layers name Layer 1 (Blue), Layer 2
(Green), Layer 3 (Red), Layer 4 (Near IR), Layer 5 (Mid-IR), Layer 7 (Mid-IR), Layer 8
(DEM), Layer 8 (Slope)
Click > File> Save Project > Test.dpr
Subset Selection
Normally, image files are large in size and difficult to process. So we will be working with
a smaller area to manage easily, which will take less memory and time. You can crop
your image on the fly in the viewer by using Subset option without changing your original
image file. You can create a "subset selection" when you start a project or during
modification.
To open the Subset Selection dialog box, do the following:
After importing image layers press the Subset Selection button.
21
Click on the image and Drag to select a subset area in the image viewer.
Alternatively, you may enter the subset coordinates. You can modify the coordinates by
typing.
Confirm with OK to return to the superordinate dialog box.

You can clear the subset selection by Clicking Clear Subset in the superordinate
dialog box.
Insert Thematic Layer

Geographic representations are organized in a series of data themes, which are known as
thematic layers. During the image classification with eCognition, you can insert shape file
as a thematic layer and you can also use it in the process of image classification (if
required).
During the new project creating or modifying time, Shape files or other vector files can be
inserted to viewer. To insert a thematic layer, do the followings:
Click the Insert button
22
Choose Thematic Layers > Insert on the menu bar of the dialog box.
Right-Click inside the thematic layer list and choose Insert from the context menu.
The Import Thematic Layer dialog box opens, which is similar to the Import Image
Layers dialog box.
Modify a Project
Using Modify a Project you can add/remove more image or thematic layer or you can
rename project. Modify a selected project by exchanging or renaming image layers or
through other operations.
To modify a project, do the following
Open a project and choose File > Modify Open Project on the main menu bar.
The Modify Project dialog box opens.
Modify the necessary things

Click OK to modify the project
Save a Project
Save the currently open project to a project file (extension .dpr).
To save a project, do the following:
Choose File > Save Project on the main menu bar.
Choose File > Save Project As on the main menu bar. The Save Project dialog
box opens. Select a folder and enter a name for the project file (.dpr). Click the Save
button to store the file.
23
4.2
Image Objects by Segmentation
The fundamental step of any eCognition

image analysis is to do segmentation of a
scene representing an imageinto image
object primitives. Thus, initial segmentation is
the subdivision of an image into separated
regions represented by basic unclassified
image objects called image object primitives.
For successful and accurate image analysis,
defining object primitives of suitable size and
shape is of utmost importance. As a rule of
thumb, good object primitives are as large as
possible, yet small enough to be used as
building blocks for the objects to be detected in
the image. Pixel is the smallest possible building block of an image, however it has mixture
of information. To get larger building blocks,
different segmentation methods are available to
form contiguous clusters of pixels that have
larger property space.
Commonly, in image processing, segmentation
is the subdivision of a digital image into smaller
partitions according to given criteria. Different
to this, within the eCognition technology,
each operation that creates new image objects
is called segmentation, no matter if the change
is achieved by subdividing or by merging
existing objects. Different segmentation
algorithms provide several methods of creating
Quad Tree Based Segmentations
of image object primitives.

The new image objects created by
segmentation are stored in a new image
object level. Each image object is defined by
a contiguous set of pixels, where each pixel
belongs to exactly one image object. Each of
the subsequent image object related
operations like classification, reshaping, resegmentation, and information extraction is
done within an image object level. Simply
said, image object levels serve as internal
working areas of the image analysis.
24
Multiresolution Segmentations
4.3
Classification of Land Cover Using Landsat ETM+ Image
Image Classification is a process of sorting pixels into a number of data categories based
on their data file values and reducing images to information classes. Similar features will
have similar spectral responses. The spectral response of a feature is unique with respect
to all other features of interest. If we quantify the spectral response of a known feature in
an image, we can use this information to find all occurrences of that feature throughout the
image.
Display the Image or Edit the Image Layer Mixing

Display the Image or Edit the Image Layer Mixing is one kind of band
combination process. Often an image contains valuable information about vegetation or
land features that is not easily visible until viewed in the right way. For this reason, in
eCognition, you have to use Display the Image or Edit the Image Layer Mixing.
The most fundamental of these techniques is to change the arrangement of the bands of
light used to make the image display. In order to display an image in eCognition,
assigns one or RGB color to each of up to three bands of reflected visible or non-visible
light.
You have the possibility to change the display of the loaded data using the Edit Layer
Mixing dialog box. This enables you to display the individual channels of a
combination.
To open the Edit Image Layer Mixing, do one of the following:
From the View menu, select Image Layer Mixing
Click View > Image Layer Mixing on the main menu bar.
Or Click on the Edit Image Layer Mixing button in the View Settings toolbar.
25
Figure 7: Edit Image Layer Mixing dialog box. Changing the layer mixing and equalizing options affects the
display of the image only
Choose a layer mixing preset:
26
(Clear): All assignments and weighting are removed from the Image Layer table
One Layer Gray displays one image layer in grayscale mode with the red,
green and blue together
False Color (Hot Metal) is recommended for single image layers with large
intensity ranges to display in a color range from black over red to white. Use this
preset for image data created with positron emission tomography (PET)
False Color (Rainbow) is recommended for single image layers to display a

visualization in rainbow colors. Here, the regular color range is converted to a
color range between blue for darker pixel intensity values and red for brighter
pixel intensity values
Three Layer Mix displays layer one in the red channel, layer two in green and
layer three in blue
Six Layer Mix displays additional layers.

For current exercise change the band combinations (B7, B2, and B1) and
Equalizing Histrogram any others
Click> OK
Create Image Objects

The fundamental step of any eCognition image analysis is a segmentation of a scene
representing an imageinto image objects. Thus, initial segmentation is the subdivision of
an image into separated regions represented by basic unclassified image objects called
Image Object Primitives.
View Settings Toolbar
The View Settings Toolbar buttons, numbered from one to four, allow you to switch
between the four window layouts. These are Load and Manage Data, Configure Analysis,
Review Results and Develop Rule Sets. As much of the User Guide centers around writing
rule sets which organize and modify image analysis algorithms the view activated by
button number four, Develop Rule Sets, is most commonly used
In the View Settings toolbar there are 4 predefined View Settings available, specific
to the different phases of a Rule Set development workflow.
View Settings toolbar with the 4 predefined View Setting buttons: Load and Manage Data,
Configure Analysis, Review Results, Develop Rule Sets.
Select the predefined View Setting number 4 Develop Rulesets from the View
Settings toolbar.
For the Develop Rulesets view, per default one viewer window for the image data is
open, as well as the Process Tree and the Image Object Information window,
the Feature View and the Class Hierarchy
27
Insert Rule for Object Creation

This is the first step of image classification in eCognition. This is a kind of assigning
condition/s. Based on this, it will create image objects or segments. Within the rule sets,
you can use variables in different ways. While developing rule sets, you commonly use
scene and object variables for storing your dedicated fine-tuning tools for reuse within
similar projects.
Insert a Process
Insert a Parent Process
A parent process is used for grouping child processes together in a hierarchy level. The
typical algorithm of the parent process is "Execute child process".
To open the Process Tree window Click Process> Process Tree
Go to the Process Tree window, which might be empty since you did not put any
process yet.
Insert a Segmentation Parent Process
Right-Click in the Process Tree window and select Append New from the
context menu.
New Dialog box (Edit Process) will be appeared.

28
In the Name field enter the name Segmentation and confirm with OK. It will be
your Parents of Segmentation.
Insert a Child Process ( Multiresolution Segmentation)

The child processes algorithm in conjunction with the no image object domain to
structure to your process tree. A process with this setting serves as a container for a
sequence of function related processes.
The first crucial decision you have to make is which algorithm to be used for creating
objects. The initial objects you create will be the basis for all further analysis.
Multiresolution Segmentation creates groups of areas of similar pixel values into
objects. Consequently, homogeneous areas result in larger objects, heterogeneous areas
in smaller ones.
Select the inserted Segmentation Process and Right-Click on it. Choose Insert
Child form the context menu.
Click Algorithm > Select Multiresolution Segmentations

Give the level name (Level-1)
29
Change the image layer weights

Change the scale parameter and etc.
Click > OK
Which layers to be used for creating Objects?
The basis of creating image objects is the input-data. According to the data and the
algorithm you use, objects results in different shapes. The first thing you have to evaluate,
which layers contain the important information. For example, we have two types of image
data, the Image and the DEM. In most Segmentation algorithms you can choose whether
you want to use all data available or only specific layer. It depends on where the
important information is contained. In our case, we want to use VIS and NIR band for
image object creation.
Which Scale Parameter to be set?
The Scale parameter is an abstract term. It is the restricting parameter to stop the objects
from getting too heterogeneous. For the Scale parameter there is no definite rule, you
have to use trial and error to find out which Scale parameter results in the objects is
useful for your further classification.
30
Right-Click one the process and select execute to execute the Multiresolution
Segmentation process.
Create Relational Feature

To open the Relational Feature window, Click Tools> Feature View
Feature View will be appeared.

Double-Click on Create new Arithmetic Feature, Edit Customize Feature will
be appeared
Assign the Feature
name > NDVI
The Normalized Difference Vegetation Index (NDVI) is a simple numerical

indicator that can be used to analyze remote sensing measurements. NDVI is related to
vegetation, where healthy vegetation reflects very well in the near infrared part of the
spectrum. Index values can range from -1.0 to 1.0, but vegetation values typically range
between 0.1 and 0.7.
Free standing water (ocean, sea, lake, river, etc.) gives a rather low reflectance in both
spectral bands and thus result in very low positive or even slightly negative NDVI values.
Soils which generally exhibit a near-infrared spectral reflectance somewhat larger than the
red, and thus tend to also generate rather small positive NDVI values (say 0.1 to 0.2).
NDVI (ETM+) = (Band 4 - Band 3) / (Band 4 + Band 3)
Double-Click on Layer Values and then Mean Layer appear
Double-Click on Landsat ETM+ band and complete the formula for NDVI
For NDVI = ([Mean Layer 4 (Near IR)]-[Mean Layer 3 (Red)])/([Mean Layer 4 (Near
IR)]+[Mean Layer 3 (Red)])
Click > OK
31
Land & Water Mask (LWM)

Land and Water Mask index is a very useful tool to differentiate between land and water.
This is very important variable to classify all type of waterbodies. Index values can range
from 0 to 255, but water values typically range between 0 and 50.
Water Mask = (MIR) / (Green) * 100
Assign the Feature name > Land & Water Mask
Land & Water Mask (LWM) = [Mean Layer 5 (Mid-IR)]/([Mean Layer 2 (Green)])*100
32
Click > OK
Insert the Class/Class Hierarchy
New Dialog box will be appear

On the Class Hierarchy Right-Click and Choose Insert Class form the context
menu and Class description dialog Box will be appeared,
On the Class description, give the class name Deep To Medium Deep Perennial Natural
Waterbodies and Click > OK
Insert a Classification Parent Process

Right-Click in the Process Tree window and select Append New from the
context menu.
33
New Dialog box will be appeared.

In the Name field enter the name Classification and confirm with OK. It will be
your parents of Classification
Select the inserted Classification Process and Right-Click on it. Choose Insert
In the Name field, enter the name Perennial Natural Waterbodies and
confirm with OK. It will be your Parents Class for a particular class (in this case, for
Deep to Medium Perennial Natural Waterbodies Class).
Assign Class Algorithm

The Assign Class algorithm is the most simple classification algorithm. It determines by
means of a threshold condition whether the image object is a member of the class or not.
34
This algorithm is used when one threshold condition is sufficient to assign an Image
Object to a Class.
Classify the Deep To Medium Deep Perennial Natural Waterbodies

Select the inserted Classification Process and Right-Click on it. Choose Insert Child
form the context menu and Assign Class Algorithm
In the Edit Process dialog box, select assign class from the Algorithm list.
35
In the algorithm parameter Use class, select Deep To Medium Deep Perennial Natural
Waterbodies.
In the Image Object Domain group Click > Select image object level
In the Image Object Domain group set the Parameter Click on Level> Select Level-1
36
In the Class Filter dialog box, Select unclassified from the classification list.
In the Image Object Domain (Parameter) group Click the Threshold

condition; it is labeled if condition is not selected yet.
From the Select Single Feature boxs Double-Click on Land & Water Mask (LWM)
assign the threshold <= 20 Click > OK to apply your settings
37
Right-Click one the process and select execute to execute the Perennial Natural
Waterbodies process or Using F5 Execute the Process.
Classify the Lake

38
In the Name field, enter the name Lake and confirm with OK
Assign Class Algorithm for Lake

Object to a Class.
39
Classify the Lake

In the algorithm parameter Use class, select Lake.
40
41

Right-Click one the process and select execute to execute the Lake process or Using
F5 Execute the Process.
42
*Note: Based on the LWM algorithm others land cover area has been classified as Lake.
So you have to use few more conditions for refining the Lake area.
In the Edit Process dialog box, select merge region from the Algorithm list and
Fusion super objects Yes
In the Image Object Domain Select Level-1 and In the Class filter Select > Lake >
OK
Using F5 Execute the algorithm

In the Edit Process dialog box, select assign class from the Algorithm list and Use
class unclassified
43
In the Image Object Domain select image object level and parameter Level >
Level-1, Class> Lake
In the parameter Click on Threshold condition and to apply your bellow settings
Feature select Area and Threshold <= 3600000

Using F5 Execute the Lake algorithm
44
Figure Classified lake area
45
Classify the River

In the Name field, enter the name River and confirm with OK
Assign Class Algorithm for River

Object to a Class.
46
Classify the River

In the algorithm parameter Use class, select River.
47
48

Right-Click one the process and select execute to execute the River process or Using
F5 Execute the Process.
49
*Note: Based on the LWM algorithm others land cover area has been classified as River.
So you have to use few more conditions for refining the
River area.
In the Edit Process dialog box, select assign class from the Algorithm list and Use
class unclassified
In the Image Object Domain select image object level and parameter Level >
Level-1, Class> River
In the parameter Click on Threshold condition and to apply your bellow settings
50
Feature select Length/Area and Threshold <= 1.6

Similar way add following condition for river and Using F5 Execute the Lake
algorithm
Classify the Broadleaved Tree Crop

Select the inserted Classification Process and right-click on it. Choose Insert Child
form the context menu.
51
New Dialog box will be appear

In the Name field enter the name Broadleaved Tree Crop and confirm with OK.
It will be your parents of Classification
Classify the Broadleaved Tree Crop

52
In the algorithm parameter Use class, select Broadleaved Tree Crop.
53

From the Select Single Feature boxs Double-Click on NDVI assign the threshold =>
0.35 Click > OK to apply your settings
54
Right-Click one the process and select execute to execute the Broadleaved Tree
Crop process or Using F5 Execute the Process.
*Note: Based on the LWM algorithm others land cover area has been classified as
Broadleaved Tree Crop. So you have to use few more conditions for refining the
Broadleaved Tree Crop area.
Similar way add other condition for Broadleaved Tree Crop and Using F5 Execute
the Broadleaved Tree Crop algorithm
55
Please set following condition for others land cover

Please note classification criteria in the examples below are simplified for the sake of the
exercise and normally would need to be more elaborate.
Bare Soil in seasonally flooded area

SLAVI stands for: Specific leaf area vegetation index
Bare Soil
Urban and Industrial Areas
Irrigated Herbaceous Crop
Rainfed Herbaceous Crop
56
Closed to Open Rooted Forb
Closed to Open Grassland
Small Herbaceous Crops in sloping land
Closed to Open Seasonally Flooded Shrubs
Closed to Open Shrubland
Small Sized Field Of Tree Crop
57
Broadleaved Tree Crop
Broadleaved Open Forest
Broadleaved Closed Forest
Classified Land cover

58
*Note The entire classification process shown base on single variable. For better results
more variable need to use.
59
4.4
Manual Editing
Manual editing of image objects and thematic objects allows you to manually influence the
result of an image analysis. The main manual editing tools are Merge Objects
Manually, Classify Image Objects Manually and Cut an Object Manually.
While manual editing is not commonly used in automated image analysis, it can be
applied to highlight or reclassify certain objects or to quickly improve the analysis result
without adjusting the applied rule set.
To open the Manual Editing toolbar choose View > Toolbars > Manual Editing on the
main menu.
Change Editing Mode

The Change Editing Mode drop-down list on the Manual Editing toolbar is set to
Image Object Editing by default. If you work with thematic layers and want to edit
them by hand, choose Thematic editing from the drop-down list.
Selection Tools
Objects to be fused or classified can be selected from the Manual Editing toolbar in
one of the following ways:
1 Single Selection Mode selects one object. Select the object with a single click.
2 Polygon Selection selects all objects that lie within or touch the border of a polygon.
Set vertices of the polygon with a single click. Right-click and choose Close Polygon to
close the polygon.
3 Line Selection selects all objects along a line. Set vertices of the line with a single
click. A line can also be closed to form a polygon by right-clicking and choosing Close
Polygon. All objects that touch the line are selected.
60
Rectangle Selection selects all objects within or touching the border of a rectangle.
Drag a rectangle to select the image objects.
Merge Objects Manually

The manual editing tool Merge Objects is used to manually merge selected
neighboring image or thematic objects.
Note: Manual object merging operates only on the current image object level.
Tools > Manual Editing > Merge Objects from the main menu bar or press the
Merge Objects Manually button on the Manual Editing toolbar to activate the input
mode. Or you can use right click.
Note: You should have at list two objects.
61
62
Classify Image Objects Manually
The manual editing tool Classify Image Objects allows easy class assignment of
selected image objects.
Manual image object classification can be used for the following purposes:
Manual correction of previous classification results including classification of previously
unclassified objects.
Classification without rule sets (in case the creation of an appropriate rule set is more
time-consuming), using the initial segmentation run for automated digitizing. Precondition:
To classify image objects manually, the project has to contain at least one image object
level and one class in the Class Hierarchy.
To perform a manual classification, do one of the following:
Choose Tools > Manual Editing > Classify Image Objects from the menu bar.
Click the Classify Image Objects button on the Manual Editing toolbar to activate
the manual classification input mode.
In the Select Class for Manual Classification dropdown list box, select the class to which you want to manually
assign objects. Note that selecting a class in the Legend
window or in the Class Hierarchy window (if available)
will not determine the class for manual editing; the class has
to be selected from the before-mentioned drop-down list.
Now objects can be classified manually with a single mouseclick. To classify objects, do one of the following:
Select the Classify Image Objects button and the Class for Manual
Classification. Click the image objects to be classified.
Select the image object(s) you want to classify first. Select the Class for Manual
Classification and press the Classify Image Objects button to classify all selected
objects.
Select one or more image objects, right-click into the image object(s) and select
Classify Selection from the context menu.
When the object is classified, it is painted in the color of the respective class.
63
If no class is selected, a mouse-click deletes the previous class assignment; the image
object becomes unclassified.
To undo a manual classification on a previously unclassified object, simply click the object
a second time. If the object was previously classified, then clicking again does not restore
the former classification; instead, the object becomes unclassified.
Export Results
To export results, open the Export Results dialog box by choosing Export > Export
Results from the main menu bar.
Choose Shapefile/Raster from the Export Type drop-down list.

From the Content Type drop-down list, choose to export shape file for: Classes
64
The Format has to be *.shp.

Select the image object Level for which you want to export results: Level-1.
Change the default file name in the Export File Name text field if desired.
To save the shape file to disk, press Export.
Note: Definiens Trial version cannot export the results
Segmentation Creates a New Image Object Level
The new image objects created by segmentation are stored in what is called an new
image object level. Each image object is defined by a contiguous set of pixels, where
each pixel belongs to exactly one image object. Each of the
subsequent image object related operations like classification,
reshaping, re-segmentation, and information extraction is done within
an image object level. Simply said, image object levels serve as
internal working areas of the image analysis.
Delete Image Object Level

Delete an image object level. This enables you to work with image object levels that are
temporary, or that might be required for testing processes while developing rule sets.
To delete an image object level do the following:
Choose Image Objects > Delete Levels on
the main menu bar.
The opening Delete Level dialog box displays a

lists of all image object levels according to the image
object hierarchy.
Select the image object level to be deleted

Confirm with OK. The selected image object levels will be removed from the image
object hierarchy.
65
4.5
Annex I: Information on Landsat and Indices
Satellite
Sensor
Band
Resolution
Landsat
ETM+
Band 1 0.45 - 0.52m (Blue)
30 meter
Band 2 0.52 - 0.60m (Green)
30 meter
Band 3 0.63 - 0.69m (Red)
30 meter
Band 4 0.75 - 0.90m (NIR)
30 meter
Band 5 1.55 - 1.75m (IR)
30 meter
Band 6 10.4 - 12.50m (TIR)
60 meter
Band 7 2.08 - 2.35m (NIR)
30 meter
Band 8 0.52 - 0.90m (Pan)
15 meter
Normalized Difference Vegetative Index (NDVI)

(ETM+) NDVI = (Band 4 - Band 3) / (Band 4 + Band 3)
Normalized Difference Snow/Ice Index (NDSII)

NDSII = (green infra-red) / (green + infra-red)
(ETM+) NDSII = (Band 2 - Band 5) / (Band 2 + Band 5)
Land and Water Masks (LWM)

Water Mask = (infra-red) / (green + .0001) * 100
(ETM+) Water Mask = (Band 5) / (Band 2 + .0001) * 100
Modification of Normalized Difference Water Index (NDWI)

NDWI=(NIR IR / (NIR + IR)
(ETM+) NDWI = (Band 4 - Band 5) / (Band 4 + Band 5)
66
Normalized Burn Ratio

NBR=(NIR TIR) / (NIR + TIR)
(ETM+) NBR = (Band 4 - Band 7) / (Band 4 + Band 7)
Ratio vegetation index

RVI=NIR / red
(ETM+) RVI = Band 4 / Band 3
Green normalized difference vegetation index

GNDVI=(NIR - Green) (NIR + Green)
(ETM+) GNDVI = (Band 4 - Band 2) / (Band 4 + Band 2)
Specific leaf area vegetation index (SLAVI)

SLAVI = NIR (Red + infra-red)
(ETM+) SLAVI = Band 4 / (Band 3 + Band 5)
Normalized Difference Moisture Index (NDMI)

NDMI = (NIR-IR)/ (NIR+IR)
(ETM+) NDMI = (Band 4 - Band 5) / (Band 4 + Band 5)
67
5. References and Useful links

User Guide eCognition Developer www.ecognition.com/
Global Land Cover Facility http://www.glcf.umd.edu/index.shtml
Global SRTM Datasets http://srtm.csi.cgiar.org/
International Centre for Integrated Mountain Development http://www.icimod.org
Mountain Geoportal http://geoportal.icimod.org
Satellite imagery http://edcsns17.cr.usgs.gov/NewEarthExplorer/
Satellite imagery http://glovis.usgs.gov/
Bajracharya, B., Uddin, K., Chettri , N., and Shrestha, B., Siddiqui , S., (2010)
Understanding Land Cover Change Using a Harmonized Classification
System in the Himalaya A Case Study From Sagarmatha National Park,
Nepal' Mountain Research and Development 30(2):143-156.
Cihlar, J. (2000). Land-cover mapping of large areas from satellites: Status and
research priorities. International Journal of Remote Sensing, 21, 10931114.
Di Gregorio, A (2005) Land Cover Classification System (LCCS), version 2:
Classification Concepts and User Manual. FAO Environment and Natural
Resources Service Series, No. 8, Rome: FAO
Franklin, S. E., & Wulder, M. A. (2002). Remote sensing methods in medium
spatial resolution satellite data land cover classification of large areas.
Progress in Physical Geography, 26, 173205.
Houghton, R.A., (1998). Historic role of forests in the global carbon cycle. In:
G.H. Kohlmaier, M. Weber, and R.A. Houghton, eds. Carbon dioxide
mitigation in forestry and wood industry. Berlin: Springer, 1 24.
John R. Townshend, Jeffrey G. Masek, Chengquan Huang, Eric F. Vermote, Feng
Gao, Saurabh Channan, Joseph O. Sexton, Min Feng, Raghuram
Narasimhan, Dohyung Kim, Kuan Song, Danxia Song, Xiao-Peng Song,
Praveen Noojipady, Bin Tan, Matthew C. Hansen, Mengxue Li & Robert E.
Wolfe (2012): Global characterization and monitoring of forest cover using
Landsat data: opportunities and challenges, International Journal of Digital
Earth, 5:5, 373-397
Pandey, D.N., (2002). Sustainability science for tropical forests. Conservation
Ecology, 6, 13.
Strategic Plan for the Climate Change Science Program Final Report, July 2003
68
Vogelmann, J. E., Howard, S. M., Yang, L. M., Larson, C. R., Wylie, B. K., & Van
Driel, N. (2004). Completion of the 1990s National Land Cover Data set for
the conterminous United States from Landsat Thematic Mapper data and
ancillary data sources. Photogrammetric Engineering and Remote Sensing,
67, 650662.
69

Land Cover Mapping Tutorial

Încărcat de

Informații document

Drepturi de autor

Formate disponibile

Partajați acest document

Partajați sau inserați document

Opțiuni de partajare

Vi se pare util acest document?

Este necorespunzător acest conținut?

Drepturi de autor:

Formate disponibile

Land Cover Mapping Tutorial

Încărcat de

Drepturi de autor:

Formate disponibile

Training Manual

Land cover mapping

Legend development and classification scheme .............................................. 3

2. Fundamental concepts in satellite image classification ............... 6

Image ratios for information extraction .......................................................... 7

3. Methodology of image analysis for land cover mapping ........... 8

Image rectification ...................................................................................... 8

Image classification ..................................................................................... 9

Object-based image analysis (OBIA) ........................................................... 10

Accuracy assessment ................................................................................. 12

Land cover change analysis ....................................................................... 13

4. Hands on Exercise Using eCognition Developer ....................... 18

Getting started .......................................................................................... 18

Create a New Project.......................................................................................... 19

Image Objects by Segmentation ................................................................. 24

Classification of Land Cover Using Landsat ETM+ Image ............................... 25

View Settings Toolbar.......................................................................................... 27

Manual Editing ......................................................................................... 60

Annex I: Information on Landsat and Indices ................................................ 66

5. References and Useful links ...................................................... 68

Land cover mapping using satellite data

Land cover mapping using satellite data

Land cover mapping using satellite data

Land cover mapping using satellite data

Legend development and classification scheme

Land cover mapping using satellite data

Table 1: legend of land cover for different levels

1.2 Available satellite images and their spatial and spectral

Land cover mapping using satellite data

Very high-resolution with 8 Band

Very high-resolution with 8 Band

High medium resolution

Medium-resolution with 8 Band

Medium-resolution with 7 Band

* Land cover map legends at different scales

Land cover mapping using satellite data

2. Fundamental concepts in satellite image classification

Land cover mapping using satellite data

Image ratios for information extraction

Normalized difference vegetation index (NDVI)

Figure 2: Satellite image and NDVI of

Figure 1: Normalized difference vegetation index of ETM+ image

NDVI index values can range fro

Normalized Difference Snow Index (NDSI)

Land cover mapping using satellite data

3. Methodology of image analysis for land cover mapping

Image classification scheme

Land cover mapping using satellite data

Land cover mapping using satellite data

Association: Association takes into account the relationship between other

Object-based image analysis (OBIA)

Classification is producing meaningful material distribution maps via identification of

Land cover mapping using satellite data

merged (Zhou et al., 2008). Definiens (eCognition) provides segmentation on several

Figure 4: Flow diagram of Hierarchal Object based Image classification scheme

Land cover mapping using satellite data

Similar features will have similar spectral responses.

Figure 5: Class Hierarchy and respective classified objects

Land cover mapping using satellite data

categories. The outcome of accuracy assessment is usually presented in a table that

A sample accuracy assessment report

Land cover change analysis

climate-related change (both long and short term);

agricultural encroachment (farming activities)