The research topic is land cover mapping with object oriented image analysis approach. The research took urban conglomeration of Vijayawada city as one of the case study. Over the last 30 years remote sensing methods have operationally been used for environmental issues, especially in areas where only insufficient field data are available for urban development and management. Satellite images were used to perform the land cover classification using the two selected image classification approaches. Classifying the complex structures of urban morphology from high resolution remote sensing imagery comprises difficulties due to their spectral and spatial heterogeneity. This research presents a methodology allowing to derive meaningful area-wide spatial information for city development and management from high resolution imagery. The initial point is a stable segmentation for an object-oriented approach to derive a thematic land cover classification. The classification methodology – which is predominantly based on shape and neighborhood related features – will be exemplified by the extraction of urban land cover classification. Finally, the urban land cover classification is used to compute a spatial distribution of built-up densities within the city and to map homogeneous zones or structures of urban morphology. The aim apart from the information on urban morphology is the opportunity to derive indirectly standard socio-economic data for further support of city management and development. The result shows an allocation of different urban zones within the city of Vijayawada with an accuracy of 82% compared to a digitized layer based on visual classification.
Up-to-date and area-wide information management in highly dynamic urban settings is a critical endeavor for their future development. Especially in explosively growing and altering cities a lack of up-to-date data is apparent. For this purpose remote sensing offers the possibility of a fast and area-wide assessment of urban changes and developments. This research describes a work-flow from an original high resolution satellite image up to a differentiation of inner urban morphology.
The challenge of classifying urban land cover from high resolution remote sensing data arises from the spectral and spatial heterogeneity of such imagery. The frequent alternation and coexistence of built-up structures, vegetation, bare soil or water areas and the heterogeneity of the objects themselves (for example roads with cars) result in distinct spectral variation within these areas of literal homogenous land cover classes. There to the high dissimilarity of functions like industrial or residential areas as well as parks or agricultural regions causes problems in terms of an indirect inferring of land use.
In order to characterize this complex highly-structured urban environment, an object-oriented approach with shape parameters and neighborhood relations provides additional analysis potential from remote sensing apart from spectral information. A multiresolution segmentation approach was presented in this research. Segmentation concepts and their theoretical background for object-oriented approaches have been summarized. In this context a recent segmentation procedure has been applied to derive real world structures.
Object-oriented classification methods have been presented in this research. Classification approaches of highly-structured urban areas are objects of research. This thesis presents an object-oriented approach allowing to derive a simple transferable urban land cover classification. Based on an initial segmentation the urban land cover classes are derived and homogenous urban density zones are differentiated. The assumption that homogenous urban patterns or morphologies correlate with urban function as well as with social-economic parameters has been verified. In conclusion, the urban zoning is an important approach for fast monitoring and extraction of further standard parameters for development and management of large complex urban areas.
2.0 STUDY AREA, DATA AND METHODOLOGY
2.1 Study Area
The study site, Vijayawada city, known as the political capital of the State, located in the south-east of India is the third largest city of Andhra Pradesh state. Vijayawada is located on the banks of the sacred Krishna River and is bounded by the Indrakiladri Hills on the West and the Budemeru River on the North. The Northern, North-Western, and South-Western parts of the City are covered by a low range of hills, while the Central, South-Western and North-Western parts are covered by rich and fertile agriculture lands with three irrigation canals criss-crossing them. The discovery of pre-historic remains belonging to the Stone Age along the banks of Krishna River, from Machlipatnam to Nagarjuna Sagar, proves that this part of the river valley had human settlements even during the stone age of Indian history.
The other details of Vijayawada city are:
Coordinates: 16.30° N 80.37° E
Area: 58 km²
Elevation: 125 m
Time zone: IST (UTC+5:30)
Population (2006): 1,025,436
High resolution multispectral IRS P-5 images were taken. This satellite carries two PAN sensors with 2.5m resolution and fore-aft stereo capability. The payload is designed to cater to applications in cartography, terrain modeling, cadastral mapping etc., These images were supplied by NRSA, Hyderabad, India.(http://www.nrsa.gov.in)
Here, the description about the land cover types and their distributions of the study area is given. Except this, the remote sensing images, ground truth used in this study are described in detail and also the data preprocessing before conducting the classification is described. Methodology to perform this research is given in figure 1.
Fig 1: Methodology
3.1 Image segmentation
Using the object oriented image analysis approach to classify the image is performed in eCognition.Object oriented processing of image information is the main feature of eCognition. The first step in eCognition is always to extract image object primitives by grouping pixels. The image objects will become building blocks for subsequent classifications and each object will be treated as a whole in the classification. The segmentation rule is to create image objects as large as possible and at the same time as small as necessary. After segmentation, a great variety of information can be derived from each object for classifying the image. In comparison to a single pixel, an image object offers substantially more information.
eCognition uses a newly developed segmentation procedure, multi-resolution segmentation, which is based on the possibility to generate image object primitives in any chosen scale. Segmentation is not an aim in itself. The purpose of image analysis can be a land cover/ land use classification or the extraction of objects of interest. However, objects of interest can in many cases be of considerable heterogeneity but a segmentation procedure following a relatively general homogeneity criterion will in most cases not be able to directly extract the final areas or objects of interest.
3.2 Segmentation procedure in Recognition
Multi-resolution segmentation is a basic procedure in eCognition for object oriented image analysis. It is used here to produce image object primitives as a first step for a further classification and other processing procedures.
Multi-resolution is a bottom up region-merging technique starting with one-pixel objects. In numerous subsequent steps, smaller image objects are merged into bigger ones. Throughout this pair-wise clustering process, the underlying optimization procedure minimizes the weighted heterogeneity of resulting image objects. In each step, that pair of adjacent image objects is merged which stands for the smallest growth of the defined heterogeneity. If the smallest growth exceeds the threshold defined by the scale parameter, the process stops. Throughout the segmentation procedure, the whole image is segmented and image objects are generated based upon several adjustable criteria of homogeneity in color and shape.
3.3 Comparison of segmentation results with different scale parameters in the study area
Figures 2 and 3 show the effect of segmentation results using different segmentation parameters. Except scale difference, the other parameters that influence the segmentation result are color, shape, smoothness and compactness but these are kept constant. Figure 2 is the segmentation result with a scale parameter 5.
Comparing this segmentation result with the original image, it is found that neighbor pixels are grouped into pixel clusters-objects, and because of the low value of scale parameter, there are too many small objects. Figure 3 is the segmentation result with scale parameter 10. It is found by comparing it with figure 2 that higher scale parameter value generates larger objects.
By visual comparison, a scale parameter of 10 is selected because the segmentation result fits the information class extraction best. Based on these parameters, segmentation process is performed.
Figure 2: Segmentation result 1 with parameters of Scale 5, color 0.8, and shape 0.2, smoothness 0.9, compactness 0.1.
Figure 3: Segmentation result 2 with parameters of Scale 10, color 0.8, and shape 0.2, smoothness 0.9, compactness 0.1.
Figure 4 is the segmentation and classification result of different zones of urban region of Vijayawada city. In this research four areas of Vijayawada city have been chosen to study the urban growth and development during the years of 1990 to 2007. The four areas are Bhavanipuram towards west, Payakapuram towards north, Krishna lanka towards south and the area near Airport towards east. These areas are shown in fig 4.
Multi-resolution segmentation, in particular, extracts regions of local contrast. The algorithm uses a description of heterogeneity weighted by the size of image objects. Small areas, which deviate in tone from their surrounding, are therefore merged with the surrounding.
However, if these smaller areas are of interest, multi-resolution segmentation could be applied with a smaller scale parameter, extracting principal image objects of smaller average size. The typical result of a segmentation run is: bigger homogeneous image objects, smaller heterogeneous image objects and smaller homogeneous image objects embedded in a high contrast region.
Figure 4: segmentation and classification result of different zones of urban region of Vijayawada city.
3.4 Image classification
Classification is the process of connecting the land cover classes with the image objects. After the process of classification, each image object is assigned to a certain (or no) class. In eCognition, the classification process is an iterative process. The classification result can be improved by editing the result: defining unclassified objects with the correct classes, correcting wrongly classified objects with the correct classes, etc.
Cities have emerged as the backbone of economies all over the world, with their higher contributions to overall employment and growth. There are many factors that determine urban competitiveness, both at the national and the international levels. The interplay of structural economic changes and geo-political developments, combined with domestic economic policy changes, sectoral contributions to growth and demographic changes determine the competitiveness of urban areas in any country.
A recent Organization for Economic Co-operation and Development (OECD) report on urban competitiveness for various regions lists out factors such as policy integration, public-private co-operation, human and capital development for the success of urban areas.
The communication gap between administrative mechanisms and end-users is very evident in India, especially when it comes to transportation policies and their implementation in urban areas. The most important issues that have come to the forefront are the condition and the management of cities in India.
In this context the area near Airport is chosen for classification to study its growth during the period 1998 to 2007. Fig 5 shows the area near Airport from IRS imagery. A closer look of these areas shown in fig’s 6 and 7 reveals the growth of this area in a haphazardly manner during the period 1998 to 2007 without proper planning and management. In early 90’s there was nothing in this area except some small houses and huts. Later on some structures have come up in a haphazard manner without an approval from competitive authority.
Fig 5: Satellite imagery of area near Airport of Vijayawada city.
Though some apartment structures have come up in this area with some planning and permissions from the authorities but 60 to 70% of this area is having small houses that have come up unauthorized without proper planning and management.
The rules are in place but they are not implemented by both administrators and end-users in the strict manner that they should be. The vital link is really a two-way communication process. Governance has to be effective in order to evoke sufficient response from the citizens, even if it involves strict authoritarian measures. The priority here should be to ensure efficient, rule-based, inclusive governance rather than perpetuate vote-bank politics.
This is a necessary function of any democracy. Neither the government nor the people can independently be held accountable for the fault-lines in a nation’s functioning. Also, public-private partnerships are important for the development of adequate infrastructure facilities but the failure to regulate private operators will definitely cause many problems for the government and the public.
Fig 6: Classification result of high resolution image of area near Airport of Vijayawada city in the year 1998.
With the growing urbanization and industrialization, the challenges facing our cities have also increased. Although the government has been undertaking several initiatives to solve the various problems, the challenges are immense. Ultimately, it is not only through government’s policies and actions, but also through the committed actions and initiatives of the community and individuals, who will refuse to turn the magnitude of the problem into an excuse for inaction, that the problems of our cities can be resolved and our cities can become clean and green havens.
The existing, emerging and future problems faced by urban local-self government can be solved with constructive suggestions from eminent scholars, which can make the life of the people in the city enjoyable in all aspects. These scholars, suggests the ways a city of 21 century can be socially just, categorically sustainable, politically participatory, economically viable and really capable of meeting future needs.
Fig 7: Classification result of high resolution image of area near Airport of Vijayawada city in the year 2007.
Similarly the other areas, Bhavanipuram towards west, Payakapuram towards north, Krishna lanka towards south have been classified and analyzed and found the same situation as that of area near Airport.
4.0 CONCLUSIONS AND RECOMMENDATIONS
In any case, the best process has been the process of decentralization. This has opened up new opportunities for decision-making at the local level and for the involvement of local communities and interest groups in the decision-making process. In some cases, this has also meant a weakening of the community process in the face of formal institutions at the local level. In many cases the buffer of the bureaucracy between the elected representatives and the people has been removed giving a carte-blanche to the elected representatives to take decisions without consulting anyone. In this regard the author of this thesis asks two important questions:
1) Does decentralization give city governments more power and resources and thus capacity to act? and
2) If city government does get more capacity to act, does this actually bring benefits to urban poor groups?
Therefore, it is recommended the following points for city planning, development and management.
Benz, U.C., Hofmann, P., Willhauck, G., Lingenfelder, I., Heynen, M. (2004). Multi-resolution, object-oriented fuzzy analysis of remote sensing data for GIS – ready information. ISPRS Journal of Photogrammetry and Remote Sensing 58, pp. 239-258.
ECognition User Guide: Multiresolution Segmentation.
ECognition, 2001 Concept and Methods.
Erdas Imagine, 8.5 2001, User Guide.
Gao Yan, Pixel based and object oriented image analysis for coal fire research.
Goward. N., Huemmrich. F and Waring, H 1994, Visible-near infrared Spe- ctral reflectance of landscape components in western Oregon: Remote Sensing environment Vol. 47, pp190-203.
Kushwaha, S.P.S., Subramanian, S.K., Chennaiah, G.CH., Ramana Murthy, J., Kameswara Rao, S.V.C., Perumal, A., and behera, G., 1996, International Remote Sensing and GIS methods for sustainable rural development. International Journal of Remote Sensing, 17, 3055-3069.
Lillesand, T.M., and Kiefer, R.W., Remote Sensing and Image Interpretation, 4th ed, John Wiley and Sons, inc. USA, 2001, ISBN: 0471255157.
Mather, Paul M, 1987, Computer Processing of Remotely-Sensed Images, St Edmundsbury Press Ltd., Bury St Edmunds, Suffolk, Wiley and Sons, ISBN: 0-471-90648-4.
Tso, B, and Mather, P.M, 2001, Classification Methods for Remotely Sensed Data, Taylor and Francis Inc, ISBN: 0-415-25909-6.
Vani K., and Sanjeevi S., Fusion of ASTER image data for enhanced mapping of landcover features, Institute of Remote Sensing, and Center for Geoscience and Engineering, Anna University.
ABOUT THE AUTHOR
Deccan college of Engg. and Technology,
Dar us Salam, Near Nampally,
Hyderabad-500001, A.P., India.
The size of Mobile Mapping Market was anticipated to be $14 billion in 2020 and is predicted to reach $40…
Esri has unveiled a new massive open online course (MOOC) entitled, "Spatial Data Science: The New Frontier in Analytics."