1 CHAPTER 1 INTRODUCTION 1.1 GENERAL Preparation of maps has been one of the challenging areas in surveying. The conventional methods are to go to area directly and take measurements, level etc and plot the map. This requires large human resources and consumes time and money. Also accuracy of this map is susceptible to human and many instruments involved. But in the present days we cannot afford time. Also Map revision is traditionally a manual task especially when maps are updated on the basis of aerial images and existing map data. For this reason, maps are typically out of date. For example, it has been reported that, for a number of reasons, the revision lag-time for topographic maps from the United States Geological Survey (USGS) is more than 23 years. To overcome these types of difficulties methods of extraction of road network from satellite image plays an important role. The different methods of road extraction from satellite image includes the Automatic method (J. B. Mena 2005) and Semi-automatic method (Jun Zhou 2006) of road extraction. In automatic method the human computer interaction is very low or it is said to be done with pure mathematical algorithms by the computer, whereas in the semi-automatic method there is human interaction also along with the help of computer. In semiautomatic road extraction, a road in the image is delineated using its geometric and photometric properties with the initial positions provided by an operator. While microcomputer made their first appearance three decades ago, it is only in the last 15 years they have become "seriously useable" machines. This situation has occurred as the consequence of a series of developments which includes: faster processing facility, large capacity, high performance and relatively inexpensive hard discs; high resolution colour monitors; CD-ROM players becoming near universal; and availability of inexpensive, high quality colour output devices and colour scanners. These hardware technology changes have gone in parallel with changes in better data conversion, software for scanners, better software for image manipulation and storage, and improvements in database management system.
31
Embed
PREPARATION OF ROAD NETWORK FROM SATELLITE IMAGERY
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
1
CHAPTER 1
INTRODUCTION
1.1 GENERAL
Preparation of maps has been one of the challenging areas in surveying. The
conventional methods are to go to area directly and take measurements, level etc and
plot the map. This requires large human resources and consumes time and money.
Also accuracy of this map is susceptible to human and many instruments involved.
But in the present days we cannot afford time. Also Map revision is traditionally a
manual task especially when maps are updated on the basis of aerial images and
existing map data. For this reason, maps are typically out of date. For example, it has
been reported that, for a number of reasons, the revision lag-time for topographic
maps from the United States Geological Survey (USGS) is more than 23 years. To
overcome these types of difficulties methods of extraction of road network from
satellite image plays an important role.
The different methods of road extraction from satellite image includes the
Automatic method (J. B. Mena 2005) and Semi-automatic method (Jun Zhou 2006) of
road extraction. In automatic method the human computer interaction is very low or it
is said to be done with pure mathematical algorithms by the computer, whereas in the
semi-automatic method there is human interaction also along with the help of
computer. In semiautomatic road extraction, a road in the image is delineated using its
geometric and photometric properties with the initial positions provided by an
operator.
While microcomputer made their first appearance three decades ago, it is only
in the last 15 years they have become "seriously useable" machines. This situation has
occurred as the consequence of a series of developments which includes: faster
processing facility, large capacity, high performance and relatively inexpensive hard
discs; high resolution colour monitors; CD-ROM players becoming near universal;
and availability of inexpensive, high quality colour output devices and colour
scanners. These hardware technology changes have gone in parallel with changes in
better data conversion, software for scanners, better software for image manipulation
and storage, and improvements in database management system.
2
The innovation and development in computer, communication and software is
contributing towards the growth of information technology. The net result of these
changes is that it is now relatively easy to create, store, retrieve, and analyze large
quantities of spatial and non-spatial data of urban and transportation system. A related
change is the rapid development of spatial information technologies such as Remote
Sensing (RS), Global Positioning System (GPS) and Geographical Information
System (GIS).This made the process like road network generation, map revision,
flood mapping, urban change detection, etc. easy as compared to the conventional
methods to the same.
Road tracking methods make assumptions about road characteristics like, roads are
elongated, road surfaces are usually homogeneous, there is adequate contrast between
road and adjacent areas, roads may not be elongated at crossings, bridges, and ramps,
road surfaces may be built from various materials that cause radiometric changes,
ground objects such as trees, houses, vehicles and shadows may occlude the road
surface and may strongly influence the road appearance, road surfaces may not have
adequate contrast with adjacent areas because of road texture, lighting conditions, and
weather conditions, the resolution of satellite images can have a significant impact on
computer vision algorithms.
One problem with these systems is that such assumptions are pre-defined and
fixed whereas image road features vary considerably. Such properties cannot be
completely predicted and they constitute the main source of problems with fully
automated systems. One solution to this problem is to adopt a semiautomatic
approach that retains the ‘‘the human in the loop” where computer vision algorithms
are used to assist humans performing these tasks.
The report briefly explains the preparation of road from satellite imagery by
the semi automatic method which includes the process like geo-referencing,
mosaicing, haze reduction, noise removal, image enhancements like contrast
stretching, filtering, and edge enhancement by using software ERDAS Imagine and
extraction of selected area and digitizing done in Arc GIS. Also uses EDM for width
measurement of roads at junctions and handheld GPS for non visible roads in the
satellite image.
3
1.2 OBJECTIVES
The main objectives of the project is extraction of roads from satellite images of the
selected 16 wards of Thiruvananthapuram Corporation, preparation of road network
which include digitization of road network using GIS, identification of missing road
using hand held GPS and road width measurement using EDM.
4
CHAPTER 2
LITERATURE REVIEW
2.1 GENERAL
The Road extraction from remotely sensed imagery has been an active
research area in map preparation for over two decades. During the past 20 years, a
number of semi-automatic and automatic methods and algorithms for road extraction
have been developed. Conventional methods of road extraction usually consist of
three main steps, road finding, road tracking, and road linking. In road finding, local
properties of the image are tested and road candidates are found using certain criteria.
The detected road candidates are then traced to form road segments. The separated
road segments are finally linked to generate a road network using geometric
constraints. In semiautomatic road extraction, a road in the image is delineated using
its geometric and photometric properties with the initial positions provided by an
operator. These methods use local geometric constraints for road tracking and linking.
Because the global structure of the road network is not considered, wrong segments
are unavoidable, and occlusions such as trees, shadows, surface anomalies, and road
width change can cause the tracking to be lost.
2.2 REMOTE SENSING
Remote sensing is the science and to some extent art of obtaining information
about an object, area, or phenomenon through the analysis of data acquired by a
device that is not in contact with the object, area, or phenomenon under
investigations. This is done by sensing and recording the reflected or emitted energy
and processing, analyzing, and applying that information. The advent of Remote
Sensing through space borne and air-borne platforms and sensors has opened new
vistas for modern, scientific surveying of earth’s natural resources. Remote sensing
data is the name given to any data where information about a location is collected
remotely, i.e. from a different location, such as collecting information about the
ground surface from inside an aircraft.
5
2.3 ERDAS Imagine 8.6
ERDAS IMAGINE is an image processing software with raster graphics editor
capabilities designed by ERDAS, Inc. for geospatial applications. ERDAS IMAGINE
is aimed primarily at geospatial raster data processing and allows the user to prepare,
display and enhance digital images for mapping use in GIS or in C ADD software. It
is a toolbox allowing the user to perform numerous operations on an image and
generate an answer to specific geographical questions. By manipulating imagery data
values and positions, it is possible to see features that would not normally be visible
and to locate geo-positions of features that would otherwise be graphical. The level of
brightness or reflectance of light from the surfaces in the image can be helpful with
vegetation analysis, prospecting for minerals etc. Other usage examples include linear
feature extraction, generation of processing work flows ("spatial models" in ERDAS
IMAGINE), import/export of data for a wide variety of formats, ortho-rectification,
mosaicing of imagery, stereo and automatic feature extraction of map data from
imagery.
The digital Image Processing done in ERDAS includes:
• Preprocessing
Geometric correction
Radiometric correction
Haze reduction
Noise removal
• Image enhancement
Contrast stretching
Filtering
Edge enhancement
2.4 ArcGIS
In the highly dynamic and complex world 'information' has become a critical
resource for effective and efficient management of organisation. Information
Technology in its various forms is enabling organizations to churn raw data into
meaningful information for effective decision making. One such form of Information
Technology (IT) is Geographic Information System (GIS). It is described as: “An
organized collection of computer hardware, software, geographic data and personnel
6
designed to efficiently capture, store, update, manipulate, analyze, and display all
forms of geographically referenced information”. According to this definition, GIS
includes not only computing capability and data, but also manages the users, and
organizations within which they function and institutional relationships that govern
their management and use of information. GIS system design and implementation
planning are not a separate process. They must occur in conjunctions with one
another.
ArcGIS is a suite consisting of a group of geographic information system
(GIS) software products produced by Esri.
ArcGIS is a system for working with maps and geographic information. It is
used for: creating and using maps; compiling geographic data; analyzing mapped
information; sharing and discovering geographic information; using maps and
geographic information in a range of applications; and managing geographic
information in a database.
The system provides an infrastructure for making maps and geographic
information available throughout an organization, across a community, and openly on
the Web.
7
2.4.1 Conceptualization of GIS
Conceptually, a GIS can be envisioned as a stacked set of map layers, where
each layer is aligned or registered to all other layers. Typically, each layer will contain
a unique geographic theme or data type. The GIS database stores both the spatial data
(where something occurs) and the attribute data (characteristics of the spatial data) for
all of the features shown on each layer. These themes may include, for example,
topography, soils, land-use, cadastral (land ownership) information, or infrastructure
such as roads, Traffic Analysis Zones (TAZ), pipelines, power lines, or sewer
networks. Figure 1 gives a schematic view of geographic layer system in GIS. By
sharing mutual geography, all layers in the GIS can be combined or overlaid in any
user-specified combination.
Fig. 1 Mapping layers of GIS
2.5 GLOBAL POSITIONING SYSTEM (GPS)
Global Positioning System (GPS) has tremendous potential for better transport
Haze compensation procedure is designed to minimize the influence of path
radiance effects. One means of haze compensation in multispectral data is to observe
the radiance recorded over target areas of essentially zero reflectance. For example,
the reflectance of deep clear water is essentially zero in the near-infrared region of the
spectrum. Therefore any signal observed over such an area represents the path
radiance, and this value can be subtracted from all pixels in the band.
Noise Removal:
Image noise is any unwanted disturbance in image data that is due to
limitation in the sensing, signal digitization or data recording process. The potential
sources of noise range from periodic drift or malfunction of a detector, to electronic
interference between sensor components to intermittent "hiccups" in the data
transmission and recording sequence. Noise can either degrade or totally mask the
true radiometric information content of a digital image. The objective of noise
removal is to restore an image close an approximation of the original scene as
possible.
3.4.2 Image Enhancement
The procedures applied to image data in order to more effectively display or
record the data for subsequent visual interpretation. Normally, image enhancement
involves techniques for increasing the visual distinctions between features in a scene.
The, objective is to create a new” images from the original image data in order to
increase the amount of information that can be visually interpreted from the data. The
enhanced images can be displayed interactively on a monitor or they can be recorded
in a hardcopy format, either in black and white or in color. There are no simple rules
for producing the single “best" image for a particular application. Often several
enhancements made from the same “raw” image are necessary.
The various image enhancements done to the imagery in ERDAS IMAGINE
includes:
15
Contrast Stretching:
Contrast stretching (often called normalization) is a simple image
enhancement technique that attempts to improve the contrast in an image by
`stretching' the range of intensity values it contains to span a desired range of values,
e.g. the full range of pixel values that the image type concerned allows. It differs from
the more sophisticated histogram equalization in that it can only apply a linear scaling
function to the image pixel values. As a result the `enhancement' is less harsh.
The intent of contrast stretching is to expand the narrow range of brightness
values typically present in an input image over a wider range of grey values. The
result is an output image that is designed to accentuate the contrast between features
of interest to the image analyst.
Contrast Stretching is to be done such that the required features will be more
clearly visible in the satellite images. The breakpoint of each band of the image is to
be adjusted in the ERDAS IMAGINE so that roads are more clearly visible. For each
band of multi-spectral images the breakpoints are to be adjusted and check whether
the roads are visible. The resultant image will give a better idea of location of roads in
the images.
Filtering:
Spatial filters emphasize or deemphasize image data of various spectral
frequencies. Spatial frequency refers to the “roughness” of the tonal variations
occurring in an image. Image areas of high spatial frequency are tonally rough. That is
gray levels in these areas change abruptly over a relatively small number of pixels
(e.g. across roads or field borders). “Smooth” image areas are those of low spatial
frequency, where gray levels vary only gradually over a relatively large number of
pixels (e.g. large agricultural fields or water bodies)
Low pass filters are designed to emphasize low frequency features (large area
changes in brightness) and deemphasize the high frequency components of an image
(local detail). A simple low pass filter may be implemented by passing a moving
window throughout an original image and creating a second image whose DN at each
pixel corresponds to the local average within the moving window at each of its
positions in the original image. Low pass filtering is done in ERDAS IMAGINE by
16
passing a moving a 3x3 pixel window throughout the original image and a low
frequency image is obtained. The low frequency image obtained after low pass filter
is smooth or blurred so that the original image details are blurred.
High pass filters do just the reverse of low pass filter. They emphasize the
detailed high frequency components of an image and deemphasize the more general
low frequency information. A simple high pass filter may be implemented by
subtracting a low pass filtered image (pixel by pixel) from the original, unprocessed
image. The high frequency image obtained after high pass filtering will have a high
contrast and gives a better idea of roads. The image will be sharpened and it roads
will be more clear.
Edge Enhancement:
In Edge enhancement it enhances the edge contrast of an image. It is typically
implemented in three steps:
• A high frequency component image is produced containing the edge
information. The Kernel size used to produce this image is chosen
based on the roughness of the image. “Rough” image suggest small
filter sizes (e.g. 3x3 pixels), whereas large sizes (9x9 pixels) are used
with “smooth” images.
• All or a fraction of the gray level in each pixel of the original scene is
added back to the high frequency component image.
• The composite image is contrast stretched. This result in an image
containing local contrast enhancement of high frequency features that
also preserves the low frequency brightness information contained in
the scene.
In ERDAS IMAGINE, the high frequency image is passed through a Kernel of
size 3x3 and a high frequency image is produced containing the edge information.
The composite image is then contrast stretched. This image is a high frequency
sharpened image. The edges of roads will be clearer in these images. This image
clearly gives the details of roads in the study area for their extraction. This road
details are then digitized in Arc GIS.
17
3.5 DIGITIZING OF EXTRACTED ROADS
The processed image is to be loaded in ArcGIS for the extraction of roads. The
roads are digitized by visual interpretation and saved as corresponding feature class
for each image. A road passing through an area with uniformly distributed vegetation,
like paddy field becomes prominent due to their different reflection characteristics.
The areas where there is a very good background contrast then the road section
throughout and edges of the road can be identified clearly.
3.6 PLOTTING OF MISSING ROADS USING GPS
Roads having smaller width are not able to digitize in ArcGIS. Those roads
can be plotted using hand held GPS. The readings, latitudes and longitudes, of roads
are to be taken manually by field investigation and need to be added to the missing
links manually.
3.7 ROAD WIDTH MEASUREMENT USING EDM
Generally the width of the road is same from junction to junction. Even though
there are slight variations but we are assuming it to be uniform. The widths of
extracted roads are to be measured using Electronic Distance Meter (EDM) at various
locations and the average value is assigned as the uniform value.
18
CHAPTER 4
DATA PROCESSING
4.1 SATELLITE IMAGES OF STUDY AREA
Cartosat images containing the selected area (Cartosat 547354 & 547355) were collected from Geo-informatics lab as it is having a high spacial resolution of 2.5m comparing to the available satellite images.
Fig. 3 Cartosat Images (2.5m)
4.2 PRE-PROCESSING OF IMAGES
In their raw form, as received from imaging sensors mounted on satellite platforms, remotely-sensed data generally contain flaws or deficiencies. The correction of deficiencies and the removal of flaws present in the data are termed pre-processing because, quite logically, such operations are carried out before the data are used for a particular purpose.
Pre-processing refers to the image rectification and restoration procedures. This is the initial step done in data processing.
Geo-referencing:
Geo-referencing of toposheet of the study area is done and the projection system adopted is Projected Coordinate System, UTM, Zone 43N. With respect to the geo-referenced toposheet the four satellite imagery were geo-referenced in ERDAS IMAGINE.
19
Fig. 4 Geo-referenced Images
Mosaicing of image:
The selected area containing the 16 wards were distributed in two Cartosat images. The Cartosat images are mosaiced to make a single image. This is done in ERDAS Imagine.
From data preparation menu by using mosaic tool the two geo-referenced Cartosat images are mosaiced to a single image.
Fig. 5 Mosaiced Image
Shapefile preparation:
From the Arc Catalog a personal Geo-database was created in that a new feature class was added with the specifications like polygonal feature, projected coordinate system as required for the shapefile. Then using the edit tool bar the boundary of the
20
selected wards is traced and saved. This is exported to ERDAS IMAGINE and the area is extracted.
Fig. 6 Shape File of the Selected Area
Extraction of selected area:
The selected area containing the 16 wards of the Trivandrum Corporation is extracted(area 25sq km) from the mosaiced Cartosat image, by preparing the shapefile of the area in ArcGIS and cutting the area from Cartosat image in ERDAS IMAGINE.
Fig.7 Extracted Image
Haze Reduction:
Haze reduction is done in ERDAS IMAGINE. The resultant images obtained after haze reduction is shown in the fig.5. For convenience haze correction routines are
21
often applied uniformly throughout a scene. The raw image will be enhanced in contrast but the image will be blurred.
Fig. 8 Haze Reduced Image
Noise Removal:
Image noise is any unwanted disturbance in image data that is due to limitation in the sensing, signal digitization or data recording process.The objective of noise removal is to restore an image close an approximation of the original scene as possible. There was not much noise in the raw data so there was not much difference in the image obtained after noise reduction.
Fig. 9 Noise Removed Image
22
4.3 IMAGE ENHANCEMENT
The procedures applied to image data in order to more effectively display or record the data for subsequent visual interpretation. Normally, image enhancement involves techniques for increasing the visual distinctions between features in a scene.
Contrast Stretching
Contrast Stretching is done such that the required features will be more clearly visible in the satellite images. The breakpoint of each band of the image is adjusted in the ERDAS IMAGINE so that roads are more clearly visible. For each band of multi-spectral images the breakpoints are adjusted and checked whether the roads are visible. The resultant image will gives a better idea of location of roads in the images.
Fig. 10 Contrast Stretched Image
High-pass Filtering
A simple high pass filter may be implemented by subtracting a low pass filtered image (pixel by pixel) from the original, unprocessed image. The high frequency image obtained after high pass filtering will have a high contrast and gives a better idea of roads. The image will be sharpened and it roads will be more clear.
23
Fig. 11 High-pass Filtered Image
Edge Enhancement
In ERDAS IMAGINE, the high frequency image is passed through a Kernel of size 3x3 and a high frequency image is produced containing the edge information. The composite image is then contrast stretched. This image is a high frequency sharpened image. The edges of roads will be clearer in these images. This image clearly gives the details of roads in the study area for their extraction. This road details are then digitized in Arc GIS.
Fig. 12 Edge Enhanced Image
24
4.4 DIGITISING OF ENHANCED IMAGES
The processed image is then loaded in ArcGIS for the extraction of roads. The roads are digitized by visual interpretation and saved as corresponding feature class for each image. A road passing through an area with uniformly distributed vegetation, like paddy field becomes prominent due to their different reflection characteristics. The areas where there is a very good background contrast then the road section throughout and edges of the road can be identified clearly. From the selected 16 wards of Trivandrum corporation 75km length of road is digitised.
Fig. 13 Digitised Road Map
25
CHAPTER 5
MAP REVISION BY FIELD DATA
The missing road networks from the satellite image due to various reasons like
the resolution of image, canopy cover, single band image, narrow width of roads etc,
are to be incorporated to the digitised road map by using collected GPS and EDM
data of the corresponding roads in ArcGIS.
5.1 GPS DATA
Roads having smaller width were not able to digitize in ArcGIS. Those roads
can be plotted using hand held GPS. The readings, latitudes and longitudes, of roads
were taken manually by field investigation and need to be added to the missing links
manually.
Table 2 GPS Coordinates of Missing Roads
LOCATION LAT LONG
Pattom 76°56´34´´ 8°31´1´´
76°56´38´´ 8°31´1´´
76°56´38´´ 8°31´5´´
76°56´38´´ 8°31´8´´
76°56´42´´ 8°31´8´´
76°56´42´´ 8°31´5´´
76°56´46´´ 8°31´5´´
76°56´46´´ 8°31´1´´
76°56´42´´ 8°31´1´´
76°56´42´´ 8°30´58´´
76°56´42´´ 8°30´54´´
26
5.2 EDM DATA
The widths of extracted roads are to be measured using Electronic Distance Meter (EDM) at various locations and the average value is assigned as the uniform value.
Table 3 Road Widths at Junctions
Junction Road Width (m)
Plamood Manchadivila 6.5
Plamood to PMG (One Way) 7
PMG to Plamood (One Way) 8
Varambasseri 5.5
Pattom 14
PMG Barton Hill 9
Museum 15
Palayam 17
Museum Vellayambalam 15
Nanthancode 7
Palayam 13.5
PMG 15
Vellayambalam Museum 15
Thiruvananthaapuram – Thenmala
15
Shasthamangalam 18
Peroorkada 14.55
27
Palayam Statue 18.5
PMG 18.5
Bakery Fly Over 21
Kerala University 23
Peroorkada Ambalamukku 13.5
Main Central 15
Kesavadasapuram Ulloor 10
Main Central 15
Pattom 14
Ayurveda College Statue 15
East Fort 15
LMS PMG 18
Palayam 14
Vellayambalam 15
Kawdiar Peroorkada 13.5
Pattom 13
Vellayambalam 13.5
Pattom Kesavadasapuram 14
Kawdiar 13
PMG 14
Medical College 7
28
Kerala University General Hospital 13
Bakery Junction 21
VJT Hall 10
PMG 6
General Hoapital Vanchiyoor 10
MG 8
Pattoor 14
Patoor Palayam-Airport 14
Kanammoola Palam 5
Pallimukku Palayam-Airport 14
Kanammoola Palam 5
Kanammoola SBI
Medical College 7
PMG 6.5
Statue Ambujavilasam 6
Press 7
29
Fig. 14 Digitized Road Network of Selected Area with GPS Data Incorporated
30
CHAPTER 6
CONCLUSION
The road map preparation using conventional methods is a tedious and time
consuming task. As the transportation facilities in the developed as well as developing
countries change at very faster rate new methods of road map preparation that make
use of the information technology is need of the time. Road extraction from satellite
image can play an important role in the map revision processes. The software like
ERDAS Imagine and Arc GIS, and Geospatial data collection instruments like GPS,
and EDM helps in the extraction of road network of an area from a satellite image
which can be used to update maps at a faster rate.
The main advantage of the approach used for the preparation of road network
using satellite imagery and other geospatial data collection mechanisms is easiness of
the work and the reduced time. Software like ERDAS Imagine saves a lot of time in
the map making process as it provides a great help in the rectification and restoration
of satellite images and further enhancement process of the image for the delineation
of the linear features like road network of an area. A geographic information system
has the power to incorporate different thematic layers of geo-spatial data and integrate
it with the non spatial data. A GIS based road network, as prepared in this work, will
facilitate further manipulation and easy updating. It can also be used for the decision
makers by employing a suitable analysis with the data.
The accuracy of the work is mainly determined by the resolution of the
satellite image used. The available high resolution image in the Department of Civil
Engineering was Cartosat image with a spatial resolution of 2.5m which was a single
band image. The results show that the width of the roads that can be extracted from
the satellite image has a relation to the spatial resolution of the data. In the present
work roads having width smaller than 5m, which is two times the spatial resolution of
the image, could not be identified in the extraction process.
Road map preparation using satellite images can eliminate a lot errors
associated with the conventional map making using field survey, especially the
inherent errors associated with the conventional plotting can be eliminated by
automatic extraction and further digitising in GIS.
31
REFERENCE
1. Ana Paula Camargo (2001). “The Uses of GPS in Civil Engineering as a Tool for Monitoring Structural Oscillations of Bridges”
2. Heng Lia, Zhen Chenb, Liang Yonga, Stephen C.W. Kongc (2004). “Application of integrated GPS and GIS technology for reducing construction waste and improving construction efficiency”
3. Jun Zhou, Walter F. Bischof Terry and Caelli (2006). “Road tracking in aerial images based on human–computer interaction and Bayesian filtering”
4. Karthika (2011). “Effect of spatial and spectral resolution on the extraction of road network”
5. Lillesand T. M., Kiefer R. W., John Wiley and sons (1979). “Remote sensing and image interpretation”
6. Mena J.B., Malpica J.A. (2005). “An automatic method for road extraction in rural and semi-urban areas starting from high resolution satellite imagery”
7. Paul M. Mather (2000). “Computer Processing of Remotely-Sensed Images An Introduction”