A Brief Overview of Digital Image Processing

Digital image processing deals with manipulation of digital images through a

digital computer. DIP focuses on developing a computer system that is able to
perform processing on an image. The input of that system is a digital image and the
system that processes the image using efficient algorithms to give an image as an
output.
Digital Image Processing is largely concerned with four basic operations: image
restoration, image enhancement, image classification, image transformation. Image
restoration is concerned with the correction and calibration of images in order to
achieve as faithful a representation of the earth surface as possible—a fundamental
consideration for all applications. Image enhancement is predominantly concerned
with the modification of images to optimize their appearance to the visual system.
Visual analysis is a key element, even in digital image processing, and the effects
of these techniques can be dramatic. Image classification refers to the computer-
assisted interpretation of images—an operation that is vital to GIS. Finally, image
transformation refers to the derivation of new imagery as a result of some
mathematical treatment of the raw image bands.
In order to undertake the operations listed in this section, it is necessary to have
access to Image Processing software such as ERDAS IMAGINE, ENVI, IDRISI,
ILWIS, ARC GIS etc. which are primarily GIS software systems, but also offer a
full suite of image processing capabilities.
Image Restoration
Remotely sensed images of the environment are typically taken at a great distance
from the earth's surface. As a result, there is a substantial atmospheric path that
electromagnetic energy must pass through before it reaches the sensor. Depending
upon the wavelengths involved and atmospheric conditions (such as particulate
matter, moisture content and turbulence), the incoming energy may be
substantially modified. Image restoration seeks to remove these degradation
effects. Broadly, image restoration can be broken down into the two sub-areas of
radiometric restoration and geometric restoration.
Radiometric Restoration
Radiometric restoration refers to the removal or diminishment of distortions in the
degree of electromagnetic energy registered by each detector. A variety of agents
can cause distortion in the values recorded for image cells. Some of the most
1
common distortions for which correction procedures exist include: uniformly
elevated values, due to atmospheric haze, which preferentially scatters short
wavelength bands (particularly the blue wavelengths); striping, due to detectors
going out of calibration; random noise, due to unpredictable and unsystematic
performance of the sensor or transmission of the data; and scan line drop out, due
to signal loss from specific detectors.
Geometric Restoration
For mapping purposes, it is essential that any form of remotely sensed imagery be
accurately registered to the proposed map base. With satellite imagery, the very
high altitude of the sensing platform results in minimal image displacements due to
relief. It is therefore necessary to use photogrammetric rectification to remove
these distortions and provide accurate map measurements.
Image Enhancement
Image enhancement is concerned with the modification of images to make them
more suited to the capabilities of human vision. Regardless of the extent of digital
intervention, visual analysis invariably plays a very strong role in all aspects of
remote sensing. Some of the image enhancement techniques are Contrast Stretch,
Composite Generation and Digital Filtering Digital sensors
Image Classification
Image classification refers to the computer-assisted interpretation of remotely
sensed images. The procedures involved are treated in detail in the IDRISI Guide
to GIS and Image Processing Volume 2 chapter Classification of Remotely Sensed
Imagery. This section provides a brief overview. Although some procedures are
able to incorporate information about such image characteristics as texture and
context, the majority of image classification is based solely on the detection of the
spectral signatures (i.e., spectral response patterns) of land cover classes. The
success with which this can be done will depend on two things: 1) the presence of
distinctive signatures for the land cover classes of interest in the band set being
used; and 2) the ability to reliably distinguish these signatures from other spectral
response patterns that may be present. There are two general approaches to image
classification: supervised and unsupervised. They differ in how the classification is
performed. In the case of supervised classification, the software system delineates
specific landcover types based on statistical characterization data drawn from
known examples in the image (known as training sites). With unsupervised

2
classification, however, clustering software is used to uncover the commonly
occurring land cover types, with the analyst providing interpretations of those
cover types at a later stage.
Supervised Classification
The first step in supervised classification is to identify examples of the information
classes (i.e., land cover types) of interest in the image. These are called training
sites. The software system is then used to develop a statistical characterization of
the reflectance for each information class. This stage is often called signature
analysis and may involve developing a characterization as simple as the mean or
the range of reflectance on each band, or as complex as detailed analyses of the
mean, variances and covariance over all bands. Once a statistical characterization
has been achieved for each information class, the image is then classified by
examining the reflectance for each pixel and making a decision about which of the
signatures it resembles most. There are several techniques for making these
decisions, called classifiers. Most Image Processing software will offer several,
based on varying decision rules. Examples of supervised classifiers include:
Parallelepiped (PIPED), Minimum Distance to Means (MINDIST), and Maximum
Likelihood (MAXLIKE). They differ only in the manner in which they develop
and use a statistical characterization of the training site data. Of the three, the
Maximum Likelihood procedure is the most sophisticated, and is unquestionably
the most widely used classifier in the classification of remotely sensed imagery.
Image Transformation
Digital Image Processing offers a limitless range of possible transformations on
remotely sensed data such as vegetation indices, PCA, color space transformations
(COLSPACE), texture calculations (TEXTURE), blackbody thermal
transformations (THERMAL), and a wide variety of adhoc transformations (such
as image ratioing) that can be most effectively accomplished with the image
calculator utility.

3
Applied Digital Image Processing and GIS in Geomorphology and
Morphometry
Morphometry referred to the measurement and mathematical analysis of the
configuration of the earth’s surface, shape and dimension of its landforms. The
morphometric analysis is done successfully through measurement of linear, aerial,
relief, gradient of channel network and contributing ground slope of the basin.
Morphometric analysis provides description of physical characteristics of the
watershed which are useful for environmental studies, such as in the areas of land
use planning, terrain elevation, soil conservation and soil erosion.
Major Morphometric Variables/Parameters
Some of the major morphometric variables analyzed using DIP and GIS include:
Aspect
Aspect generally refers to the direction to which a mountain slope faces. The
aspect of a slope can make very significant influences on its local climate because
the sun’s rays are in the west at the hottest time of day in the afternoon, and so in
most cases a west-facing slope will be warmer than sheltered east-facing slope.
The east-fasting slopes on the other hand have a higher moisture content and lower
evaporation rate. Aspect can therefore have can have major effects on the
distribution of vegetation in watersheds. Compass direction of raster data sets are
normally used to indicate aspect.
Slope
Slope analysis is an important parameter in geomorphological studies for
watershed development and important for morphometric analysis. Slopes can be
calculated from processed DEM satellite images like SRTM and ASTER DEM
data. For instance, slope maps of study areas may be calculated based on ASTER
DEM data using the spatial analysis tool in ARC GIS software. Slope grid is
identified as the maximum rate of change in value from each cell to its neighbors.
Higher slope degree results in rapid runoff and increased erosion rate (potential
soil loss) with less ground water recharge potential.
Relative relief
Relative relief is an important morphometric variable used for the assessment of
morphological characteristics of any topography. Relief is readily
calculated/determined from processed DEMS. Areas of high relief indicate high
4
runoff and decreased water availability and accessibility, while low relief indicates
flat to gentle slope types. Areas of low relief could be basically used for
agricultural activities around stream sides due to its flat nature and water
availability/accessibility.
Other morphometric variables include Stream length (Lu), stream number (the
count of stream channels in each order), Bifurcation ratio (Rb), Relief ratio (Rh),
and Length of overland flow.
The drainage morphometric analyses are prerequisite for selection of water
recharge site, watershed modeling, runoff modeling, watershed delineation,
groundwater prospect mapping and geotechnical. Quantitative morphometric
analysis of watersheds can provide information about the hydrological nature of
the rocks exposed within the watershed. A drainage map of basin provides a
reliable index of permeability of rocks and their relationship between rock type,
structures and their hydrological status. Watershed characterization and
management require detailed information for topography, drainage network, water
divide, channel length, geomorphologic and geological setup of the area for proper
watershed management and implementation plan for water conservation measures.
Digital image processing and GIS have been largely employed in providing these
information.
The remote sensing technique is conveniently employed in morphometric analysis
as satellite images provide a synoptic view of a large area and is very useful in the
analysis of drainage basin morphometry. The fast emerging spatial information
technology, remote sensing, GIS, and GPS are tools that have been used to
overcome most of the problems of land and water resources planning and
management .For instance, GIS-based evaluation using Shuttle Radar Topographic
Mission (SRTM) and Advanced Spaceborne Thermal Emission and Reflection
Radiometer (ASTER) data has given a precise, fast, and an inexpensive way for
analyzing hydrological systems. Processed DEMs have been used successfully for
generating the stream network and other supporting layers.
Digital elevation models (DEMs), such as from the Shuttle Radar Topography
Mission (SRTM), or the ASTER GDEM product (USGS, Denver, Colorado,
USA), have been used to extract different geomorphological parameters of
drainage basins, including drainage networks, catchment divides, slope gradient
and aspect, and upstream flow contributing areas. GIS based watershed evaluation
using Shuttle Radar Topographic Mission (SRTM) data has given a precise, fast,

5
and an inexpensive way for analyzing hydrological systems. Recently remote
sensing data and digital elevation models have been utilized in extracting
catchment hydrological parameters for management of water resources and the
results have shown that DEMs are very accurate tools for morphometric parameter
evaluation and watershed delineation for watershed management.
The morphometric parameters such as number of stream segments, stream order,
drainage pattern, sub-basin length, perimeter and area are often delineated from the
SRTM DEM (90m) data and verified on topographical maps. The observed
parameters are used for computation as morphometric characteristics such as
drainage density, stream frequency, bifurcation ratio, form factor, circulation ratio,
relief ratio and overland flow using mathematical equations in the GIS
environment. Moreover, the slope and relief of the area are extracted from the
same data source using spatial analyst tool in ArcGIS 10 software.
The following procedure may be followed for morphometric analysis of
watersheds.
1. Firstly, SRTM DEM will have to be downloaded from USGS and other
donor sites and primary processing will have to be performed on the image.
2. The Geocoded topographic sheets will have to be correlated with SRTM
data for delineation of watershed.
3. Watershed areas may be delineated from SRTM data along with the
topographic sheets of the study areas using GIS Software such as Erdas
Imagine, ENVI, etc.
4. All the extracted parameters from SRTM and topographical sheets, like the
number and lengths of streams of each different order, drainage area, basin
perimeter and total basin length, and width can be calculated using ARC
GIS software; drainage density, drainage frequency, shape, form factor,
circulatory ratio, and elongation ratio, were calculated from these
parameters.

Application of DIP and GIS Soil Erosion Modelling

Assessing the soil erosion rate is essential for the development of adequate erosion
prevention measures for sustainable management of land and water resources. The
amount of soil erosion is mainly affected by vegetation cover, topographic
features, climatic variables, and soil characteristics. The human activities and
large-scale developments alter the vegetation cover, impacting upon the soil

6
erosion rate. Topographic features such as ground slope, slope length, and shape
mostly affect rill and interrill erosion. In addition climatic variables such as rainfall
amount & precipitation intensity (called rainfall erosivity) and temperature
important climatic variable since it affects the vegetative materials useful in
mulching to control erosion. Soil erodibility is mainly affected by aggregate
stability, texture, depth, organic matter, and stoniness. Some of these
characteristics and features are identified and measured directly from digitally
processed images.

Satellite data can be used for studying erosional features, such as gullies, rainfall
interception by vegetation and vegetation cover factor. Utilization of multi-
temporal satellite images provides the opportunities to extract valuable information
associated to seasonal land use change. Digital Elevation Model (DEM), one of the
vital inputs required for soil erosion modelling can be created by analysis of
stereoscopic optical and microwave remote sensing data. The Remote Sensing
(RS) technology has been used to provide the land use/cover information by using
digital image processing techniques. Remote Sensing have also been employed in
soil erosion prone areas assessment and estimating soil loss and sediment yield.
Geographic Information System (GIS) technologies are valuable tools in
developing environmental models through their advance features of data storage,
management, analysis, and display.
Most soil erosion models only calculate the amount of soil erosion based on the
relationships between various erosion factors. However, digitally processed Images
and GIS integrated erosion prediction models cannot only be used to estimate soil
loss but can also provide the spatial distributions of the erosion. Especially,
generating accurate erosion risk maps in GIS environment is very important to
locate the areas with high erosion risks and to develop adequate erosion prevention
techniques. Digital image processing and GIS have been successfully used for land
degradation and erosion mapping. Studies have indicated that DIP and GIS
analysis provide satisfactory results in developing erosion surveys and risk maps
by using GIS data layers such as DEM, slope, aspect, and land use.
The most common empirical erosion prediction models are integrated with DIP
and GIS. Examples of the models are Revised Universal Soil Loss Equation
(RUSLE), The Water Erosion Prediction Project (WEPP), and COoRdination of
INformation on the Environment (CORINE). These models can be used for erosion
risk mapping. The USLE is used to estimate the annual soil loss per unit area based
on erosion factors including soil erodibility, topography, rainfall, and vegetation
cover. In the WEPP model, sediment yield and erosion rates are estimated for
multiple time periods based on specific erosion factors. In CORINE model, actual

7
soil erosion risk is determined by combining two parameters including potential
soil erosion risk data and vegetation cover data. The potential soil erosion risk is
calculated as a function of soil erodibility, erosivity, and topography. Analyzed
satellite data, soil and land cover maps and DEM derived and ancillary soil and
agro-climatic rainfall data are the basic inputs used in USLE for computation of
soil loss. Sediment Yield Prediction Equation (SYPE) also uses RS derived soil
and land use information, DEM derived slope and ancillary rainfall and
temperature data.
The vegetation cover data is very important parameter in erosion models since
intensity of vegetation cover significantly affects erosion rates. High-resolution
satellite imagery, image classification techniques have been used to generate
accurate and reliable land use/cover data. High-resolution dataset helps in
identification of erosion prone areas and provides data inputs to many of the soil
erosion models.

The use of USLE

The USLE is the most commonly used estimator of soil loss caused by overland
erosion. It was developed to predict average annual soil loss from sheet and rill
erosion, not gully or other forms of erosion. The USLE is used to

1) Predict average annual soil movement from a given field slope under
specified land use and management conditions.
2) Guide the selection of conservation practices for specific sites
3) Estimate the reduction in soil loss that would result from a change in cropping
or conservation practices
4) Determine how conservation practices may be applied or altered to allow more
intensive cultivation
Estimate soil losses from land use areas other than agricultural purposes
5) Provide soil loss estimation for determining conservation needs. The USLE is
derived empirically as

A = R × K × LS × C × P

This equation, describes soil loss (A) as a function of rainfall (R) amount and
intensity; soil erodibility (K) related to texture, percentage of organic matter,
structure, and permeability of soil; morphology, especially the length of the slope

8
(L) and the slope gradient (S); vegetation cover (C); and erosion control practices
(P).

The following flow chart shows the procedure that may be employed while soil
erosion analysis.

The five parameters of USLE can be evaluated using remotely sensed data, ground
observation and existing map data. Remotely sensed data provides the main source
of information for the establishment of land-use/land-cover, geology,
geomorphology, and soil map, as well as for deriving a scheme of watershed
distribution using both digital image processing and visual interpretation. Slope
and aspect map layers are generated from a digital elevation model. These are, in
turn, generated from elevation contours of topographic maps. Rainfall amount and
intensity maps are established based on observation data obtained over several
decades. Each of the USLE factors with associated attribute data is digitally
encoded in a GIS database to eventually produce five thematic layers which are
spatially overlaid to produce a resultant polygonal layer.

9
Applied DIP and GIS in Landscape ecology
Landscape ecology is the study of landscapes specifically; the composition,
structure and function of landscapes. It involves the study of the landscape
patterns, the interactions among the elements of these patterns, and how patterns
and interactions change over time.

Remote sensing is frequently used to generate maps of terrestrial ecosystems,

which are often based on a map that delineates different vegetation types or land
uses. Figure 1 is a simple example of an image that has been classified into five
different types of cover based on how the spectral characteristics varied across the.
One of the simplest and most common maps shows areas of forest and of non-
forest, the latter often is being a result of conversion to other uses. One point to
remember is that it is essential that such maps be calibrated and validated with
ground observations; otherwise the level of mis-classification can be very high, as
well as unknown. In addition, even with ground calibration such maps are often
only about 80 % accurate; meaning that one pixel in five is classified incorrectly.
Another point to remember is that spectrally similar vegetation types will be put in
the same class. For example, tree plantations such as oil palm may be spectrally
similar to native vegetation and both could be placed in the ‘forest’ class, and often
are.

Fig. 1 Example of a simple classification with five classes. A national park in

Thailand is in the centre of the image—a heart-shaped plateau surrounded by forest
on the downslope, and then agriculture.

10
In general, ecological research refers to the investigation of organisms and their
surrounding environment, including biotic and abiotic entities. Due to the
multifaceted nature of biodiversity, it is difficult to simply express and measure
biodiversity. Biodiversity should be related to not only the variation of life forms,
but also the ecological complexes of which they are a part. Conservation has
become an indispensable way of dealing with the accelerated native ecosystem
conversion and

Two categories of remote sensing approaches are employed in ecological studies,

namely direct and indirect remote sensing approaches.
The direct approach
The direct approach refers to the direct observation of individual organisms,
species assemblages, or ecological communities from airborne or satellite sensors,
such as the application of high spatial resolution and hyperspectral sensors.
Generally speaking, high spatial resolution, also called fine spatial resolution, is
less than 10 m, and ranges from 0.5–10 m in the commercial domain for
environmental research. IKONOS, QuickBird, OrbView-3 and SPOT-5 (Satellite
Pour l’Observation de la Terre-5) are the commonly used systems. The benefit of
high spatial resolution imagery is that it greatly increases the accuracy of
identification and characterization of small objects at spatial scales. For example,
they can be used to accurately identify plant species. High resolution IKONOS-2
satellite images was utilized to explicitly recognize the transitional areas between
tree crowns and tree shades (tree shadows), and then for the quantification of
canopy cover. Further, IKONOS imagery can be used to quantify and evaluate the
spatial structure of critical habitats and how it affects endemic species, based on
the high spatial resolution imagery. In the context of marine applications, areas of
coastline, with their fertile soil and unique flora and fauna which need to be highly
protected, can be planned for in a sustainable way through mapping the changes in
land use of the area based high resolution imagery.
Hyperspectral data have the ability to collect ample spectral information across a
continuous spectrum generally with 100 or more contiguous spectral bands. It is
different from multispectral sensors which detect relatively few discrete bands.
Hundreds of spectral bands with 10-20 nm spectral bandwidths offer new
possibilities to detect subtle differences between objects of interest. The best
example is their ability to discriminate fine-scale, species-specific land cover, such
as vegetation categories or soil types, which make remarkable contribution to the
study regarding biodiversity patterns. Hyperspectral data have been successfully
11
applied in recording information regarding critical plant properties (e.g., leaf
pigment, water content and chemical composition), discriminating tree species in
landscapes, and fairly accurate identification between different species.

Indirect approaches
Indirect approaches rely on environmental parameters derived from remotely
sensed data as proxies. For example, many species are restricted to discrete
habitats such as woodland, grassland or sea-grass beds that can be clearly
identified at a distance. By combining information about the known habitat
requirements of species with maps of land-cover derived from satellite imagery,
precise estimates of potential species ranges and patterns of species richness are
possible.

Ecological remote sensing is in three main areas. First, land cover classification,
the physiographical characteristics of the surface environment, can be used to
identify very specific habitats and predict the distribution of both individual
species and species assemblages at a large spatial extent. Secondly, integrated
ecosystem measurements offer measurements of functions at different spatial
scales, including whole ecosystems, such as the derivation of leaf area index (LAI)
and net primary productivity (NPP) mostly based on the normalized difference
vegetation index (NDVI). Thirdly, change detection provides near-continuous,
long-term measurements of key ecological parameters in order to monitor
ecosystem through time and over significant areas, such as the application of
climate change and habitat loss.

Biophysical variables can be derived from moderate resolution thermal and

multispectral remote sensing data. Such biophysical include, Enhanced Vegetation
Index (EVI), Normalized Difference Vegetation Index (NDVI). Thermal Infrared
imageries are important source of information in studying environmental
disturbance because of the negative relationship between vegetation density and
land surface temperatures. They are applied in detecting continental-scale
disturbance events such as wildfire, irrigated vegetation, precipitation variability,
and the incremental process of recovery of disturbed landscapes. Another good use
of TIR remote sensing data is to measure evapotranspiration, evaporation, and soil
moisture.
GIS Roles in Ecological studies
Information needs for biodiversity are many and varied. Any database that deals
with biodiversity information has to be geographically based, and able to predict
12
where new populations of endangered species with a limited known range might be
expected, indicating potential hot spots. An important tool for monitoring
biodiversity is a geographic information system (GIS), which accommodates large
varieties of spatial and aspatial (attribute) data. The information embedded in a
GIS is used to target surveys and monitoring schemes. Data on species and habitat
distribution from different dates allow monitoring of the location and the extent of
change.

13