1343
ANALYSIS OF DEM GENERATED USING CARTOSAT-1 STEREO DATA
OVER MAUSANNE LES ALPILES
- CARTOSAT SCIENTIFIC APPRAISAL PROGRAMME (CSAP TS - 5)
Y V N Krishna Murthÿ a , S Srinivasa Rao a , D.S. Prakasa Rao a , V. Jayaraman b
a RRSSC/ISRO, Amravati Road, Nagpur, INDIA-440033 (yvnk, ssrao, dsprakash)@isro.gov.in
b ISRO HQ, Bangalore - 500000 vjay@isro.gov.in
Commission I, SS-11
KEY WORDS: CARTOSAT-1, geometry, digital elevation model, errors, quality, RPC, Ortho-image
ABSTRACT:
The high-resolution Cartosat-1 stereo image data is expected to have significant impact in topographic mapping and watershed
applications. The objective of the present investigation is to make an assessment of the potential of the Cartosat-1 data in topographic
mapping and generation of DEMs in operational mode using COTS packages. The assessment is based on the entropy analysis of the
fore/aft images, geometric quality of the fore/aft images (2-D perspective), generation of high-resolution Digital Elevation Model (DEM),
Quality evaluation in different elevation strata, generation of ortho-image and associated accuracy assessment and application potential of
the photogrammetric products for natural resources applications. The results of the study are part of phase I of C-SAP participation as a
Co-Investigator (Col) in the evaluation of CARTOSAT-1 data for test site TS5 (Mausanne-les-Alpilles, France). The inherent orientation
accuracy of the Rational Polynomial Coefficients (RPC) models is established by comparison to the available ground truth. Ground
control points are used to correct the RPC (bias correction and also higher order transformations) in order to assess the achievable
accuracy for the generation of digital surface models (DSM). The evaluation of the geometric accuracy is based on the estimated sensor
model parameters and on the residuals at the image control and check points using various configurations of control points. The DEM
evaluation is based on the elevation differences at control points, the spatial distribution of the DEM differences and terrain gradients.
The ortho-image is evaluated with respect to available reference data. The potential of thematic content for Cartosat-1 for mapping
applications is tested based on the detection and identification of various topographic features. The average entropy of the image is
around 5.6. The standard deviation, of the CartoSAT-1 DEM in comparison with high quality reference DEM, is of the order of 3.8m to
7.3m in different topographic conditions.
1. INTRODUCTION
1.1 Study Aim
CARTOSAT-1 is designed for cartography applications. Satellite
is placed in the polar Sun Synchronous Orbit of 618 km from
Earth. CARTOSAT-1 has two high-resolution panchromatic
sensors recording stereo images in the visible radiation. The
recording swath is about 30 km with a spatial resolution of 2.5 m.
The two sensors are acquiring stereo pairs from the same orbit.
Images are obtained in fore/aft directions in near real time by the
two sensors. The products of Cartosat are more useful in
cartographic mapping and various GIS applications.
The unique characteristics
are given below.
Spatial Resolution
Radiometric Resolution
Swath
Product Dimensions
Approximate Cost
National Level DEM
Planimetric Accuracy
Vertical Accuracy
Processing Tools
of CartoSAT-1 and planned products
- 2.5m
- 1 Obits
- 30km
- 30km * 30km
- ~ 1 USS/sq.km
- CartoDEM
- 15m (CEP90)
- 8m (LE 90)
-SAPHIRE 1.0
- SST/ASST
- COTS Tools
The study objectives are:
• to analyse 2-D Geometry of Cartosat-1 stereo pairs
• to analyse Radiometry of Cartosat-1 stereo pairs
• to generate DEM from Cartosat-1 stereo pairs
• to analyse DEM wrt Reference DEM
• to generate Ortho-images using Cartosat-1 derived
DEM
• to analyse Positional Accuracy of Ortho-images
1.2 Study Area
(Test site 5 - Mausanne-les-Alpilles): The spatial extent of the
study area is 4d 39’ 47.16” to 5d 06’ 55.29” East longitude and
43d 39’ 22.90” to 43d 50’ 12.40” North latitude, falling in UTM
31 zone.
The study site is located near to Mausanne-les-Alpilles (France)
(Figure 1) and presents a low mountain terrain with elevation up
to around 650m above mean sea level, mostly covered by forest,
surrounded by low-lying agricultural plains. A number of small