CADASTRAL RESURVEY USING HIGH RESOLUTION SATELLITE ORTHO IMAGE-CHALLENGES : A CASE STUDY IN ODISHA , INDIA .

Advancements in satellite sensor technology enabling capturing of geometrically accurate images of earth’s surface coupled with DGPS/ETS and GIS technology holds the capability of large scale mapping of land resources at cadastral level. High Resolution Satellite Images depict field bunds distinctly. Thus plot parcels are to be delineated from cloud free ortho-images and obscured/difficult areas are to be surveyed using DGPS and ETS. The vector datasets thus derived through RS/DGPS/ETS survey are to be integrated in GIS environment to generate the base cadastral vector datasets for further settlement/title confirmation activities. The objective of this paper is to illustrate the efficacy of a hybrid methodology employed in Pitambarpur Sasana village under Digapahandi Tahasil of Ganjam district, as a pilot project, particularly in Odisha scenario where the land parcel size is very small. One of the significant observations of the study is matching of Cadastral map area i.e. 315.454 Acres, the image map area i.e. 314.887 Acres and RoR area i.e. 313.815 Acre. It was revealed that 79% of plots derived by high-tech survey method show acceptable level of accuracy despite the fact that the mode of area measurement by ground and automated method has significant variability. The variations are more in case of Government lands, Temple/Trust lands, Common Property Resources and plots near to river/nalas etc. The study indicates that the adopted technology can be extended to other districts and cadastral resurvey and updating work can be done for larger areas of the country using this methodology.


Introduction
Generation of Cadastral maps is possible through High-tech survey methods using Ortho-images, DGPS and ETS.Cadastral boundary vectors obtained from ortho-images are used as base maps and obscured/difficult areas (not delineated/mapped from ortho-images) are surveyed using DGPS and ETS.The vector datasets thus derived through RS/DGPS/TS survey are integrated in GIS environment to generate the base cadastral vector datasets for further settlement/title confirmation activities.This exercise is undertaken to demonstrate the efficacy of the hybrid technology for cadastral map preparation.The project was conducted in Pitambarpur Sasana village under Digapahandi Tahasil of Ganjam district, Odisha.

2.0
The objective of the project  To generate cadastral maps by using hightech survey methods using ortho-images, DGPS and ETS. To store the cadastral maps in digital format (storing & updating cadastral maps using state-of-art technology) and to use this as base for all type of revenue administration and development planning. Deriving lat-longs of cadastral maps, its mosaiking and Geo-referencing to generate Land Information System (LIS) for Tahasils.
 To use GIS applications for plot level RoR, land use and infrastructure information generation using digital cadastral database.

Methodology
The broad methodology adopted under this Pilot project is as follows: -   Maps collected in sheets were scanned after quality checking of each sheet with regard to its physical condition, readability, content and clarity.Maps were tiled on 100mx100m dimension of grid cells.All the four cadastral map sheets were digitized as per the guidelines of Department of Space.Hardcopy printouts are taken for 1 to 1 matching and quality assurance.
After necessary quality check the existing cadastral maps are converted to GIS format for further georectification with orthoimages.The detail methodology is described in Fig. 3.The pre-field map was used for ETS survey for the obscure areas where existing ground control coordinates were used for the measurement of coordinates in the obscure areas.The co-ordinates then transferred to the plot vector geo-database for the finalization of complete plot vector extraction of the mouza.The plot numbers were transferred from the geo-referenced cadastral map as well as from data collected from the tehesils for the mutated plots.ROR data for the mouza then linked with plot vector for finalization of the geodatabase.

Survey and mapping of difficult areas/ ground truth collection
DGPS observation was conducted in the field for establishment of a primary control point in the study area by a 12-hour observation.Simultaneous Observation for one hour was undertaken using a Rover for Secondary / Tertiary Control point.Further Ground control points were also retrieved by DGPS-900 with real time processing capabilities at Pitambarpur Sasan village.Field verification was also conducted for the following locations.actual field measurements.The results of these steps of accuracy checking are discussed below.

Comparison of Geo-coordinates
The geo-coordinates measured at random on the ortho image and the respective location on the ground by DGPS to check the accuracy as well as consistency based on the RMS of the photogrammetric block.
The table below depicts geo-coordinate comparison.
As may be observed, the measurements match up to second places after second decimal point.

Tie line measurement
For tie line measurement five tie lines were selected.
The length of the tie line was measured from the image in GIS environment and its corresponding field measurement was done through DGPS.The results are tabulated in table-2 & Fig- 5 for comparison.The measurements were found to be matching upto third places after decimal point.

Parcel Boundary Measurement
Five plots in study area were randomly selected.Four sides of each of these plots were measured on the ground using meter-tape.The corresponding measurements were extracted from the image.The measurements were tabulated in the table.3for comparison.The difference in measurements in each case is below 1%.

CONCLUSION
High-resolution space-borne remote sensing image data show a high level of detail and provide many opportunities to be used as base for cadastral map generation.Ortho images generated by using satellite data having 0.5 m spatial resolution are ideally suited for deriving cadastral plot vectors for plain areas.The obscured areas need ground survey intervention by DGPS & ETS.The habitation area vectors (very small polygons which cannot be resolved through 0.5 m data) of existing cadastral maps can be integrated to image vector maps to finalise the new cadastral maps of the villages.The image derived cadastral maps can be directly used by revenue official for tenant interaction, settlement activities and revenue administration.
One of the significant observations of the study is matching of Total village area in Cadastral map area (after digitization), the image map of 2009 and RoR area.
The final cadastral map generated (fig.6) by Hightech survey provides accurate matching of plot areas in 81% of plots.In total 80% of plots in both the villages derived by high-tech survey method shows acceptable level of accuracy considering the fact that the mode of area measurement by ground and automated method has significant variability.
Including the Gharabari plots the total plot area in 95% (within 0-2% variations) of plots are matching with existing RoR.
It is also observed that the plot area of digitized cadastral map and the image map are matching but in case of certain plots the RoR area shows wide variation.The variation is more in case of Government lands, Temple/Trust lands, Common Property Resources and plots near to Village boundary etc. as the tenants trespassed to the vacant Govt.lands.
The adopted technology can be successfully used for Cadastral Resurvey and Cadastral GIS generation for plain areas of the state.
Fig 6 Acquisition of digital stereo satellite data of World View-II. Establishment of GCP control network by DGPS. Generation of photogrammetric block. Ortho image generation  Collection of existing cadastral maps & Coding/Scanning/Digitization of maps  Delineation & confirmation of village boundary  Plot level vectorisation & map generation from ortho image  Integration of image derived vectors and cadastral vectors (DoLR map) / ROR linking  Survey and mapping of difficult areas/ ground truth collection  Integration of vectors and preparation of base maps  Map/RoR printing (output generation) of World View-II, PAN was acquired through National Remote Sensing Centre, Hyderabad, Department of Space, and Govt. of India having spatial resolution of .5 m for this study.
Fig 1 Fig 2 The outer village boundary along with the location of bijunction and triunction points are extracted from digital cadastral map and transferred to the orthoimage by georeferncing the cadastral map with orthoimage.The field observations were taken using DGPS to confirm the outer boundary of the village and its coexistence with the neighboring mouza.The plot boundaries / vectors are interpreted by onscreen digitization method from enhanced Worldview orthoimage with precision following segmented approach i.e. the mouza divided into zones taking natural boundary into consideration and the sum of the area of the zones matching with ROR area.The vectors are drawn for the plots, which are clearly visible on the image in the first instance.The plots whose boundaries are obscured because of tree canopy and other reasons are not closed and left as it is.These difficult area polygons are finalized by field survey technique using DGPS and ETS.The habitation polygons, which are very small, are incorporated to the map from the existing cadastral map.A pre-field map was generated in 1:2000 scales for ground validation and DGPS based data collection.


Field boundaries obscured by tree cover (Vertices closed by field measurements) Example of difficult areas is shown in Fig.4. Plot boundaries not clear on the image  Govt.land under holding (cadastral plot showing subplots on image) by individuals  Acquisition of plots (road and other infrastructure like Tank/ Canal/ Water Harvesting Structures (WHS) showing no plot boundary on image but multiple plots on cadastral map & Record of Rights (ROR)  Alteration of plot boundaries (locally) by owners  Verification of same plot in ROR  Encroachment of land (Temple/ Trust board land, canal and road side, Government land) for which cadastral plot boundary and image boundary not matching  Water channel (disused) encroached by nearby plot owners. Cadastral map showing two plots but on the image it is one plot  Cadastral map showing one plot but on the image it is two plot  Boundary plot (village boundary) truncated due to acquisition of adjacent plot of another village by same owner and subsequent removal of common boundary  Village boundary (boundary plot) vector changed after construction of Water Harvesting Structures (WHS)  Linear measurements in two villages to check the map accuracy  Sample DGPS observation on plot corner (vertices) and post plotting for accuracy checking  Ground truth on zero fill parcels, part/divide /Joint parcels were undertaken and Geocoordinates were taken up using DGPS (RTK) at few checkpoints for accuracy checking.