Manzul Kumar Hazarika
Panchromatic data, the estimated width varies from 60m to 70m with a maximum error of 6m. ADEOS Multispectral
and LANDSAT TM data can estimate such a road with a maximum error of 16m and 26m respectively.
A 35m wide road varies from 34m to 36m in SPIN-2 data, giving a maximum error of 1m. Width of the same road of
section road varies from 32m to 40m and 30m to 40m in ADEOS Panchromatic data and SPOT Panchromatic data
respectively with a maximum error of 5m in each case. ADEOS Multispectral and LANDSAT TM data give a
maximum error of 13m and 25m respectively while estimating the width.
SPIN-2 data measures a 15m wide road with a maximum error of 1m. ADEOS Panchromatic and SPOT Panchromatic
data can measure the width of the road with a maximum error of 7m and 5m respectively whereas ADEOS
Multispectral and LANDSAT TM data can not recognise such a road.
A 5m wide road can be measured with a maximum error of 1m using SPIN-2 data.
1 2 3 4 S 6 7 8 9
Actual | Satellite sensors | No. of | Pixel | Range of | Mean | Error Range Maximum
width pixels size estimated (6-1) of error error
width (5-1)
meter meter meter | meter meter meter
SPIN-2 31-32 2 62-64 63 -1 -2-0 2
ADEOS Pan 7-8 8 56-64 60 -4 -8-0 8
64m | SPOT Pan 6-7 10 60-70 65 1 -4 ~ +6 6
ADEOS Mul" 3-4 16 48-64 56 -8 -16 - 40 16
LANDSAT TM 2-3 30 60-90 75 11 -4 - +26 26
SPIN-2 17-18 2 34-36 35 0 -1 ~ +1 1
ADEOS Pan 4-5 8 32-40 36 1 -3~45
35m | SPOT Pan 3-4 10 30-40 35 0 -5 - +5 5
ADEOS Mul 2-3 16 32-48 40 S -3~+13 13
LANDSAT TM 1-2 30 30-60 45 10 -5 ~ +25 25
SPIN-2 7-8 2 14-16 15 0 -1 ~ +1 1
ADEOS Pan 1-2 8 8-16 12 -3 -7 ~ +1 7
15m | SPOT Pan 1-2 10 10-20 15 0 -5 - +5 5
ADEOS Mul - 16 - - - - =
LANDSAT TM - 30 = - - - -
SPIN-2 2-3 2 4-6 3 0 -1 ~ +1 1
ADEOS Pan = 8 = - = -
5m SPOT Pan = 10 - - - = -
ADEOS Mul - 16 - - - - -
LANDSAT TM - 30 = - - - -
"Panchromatic; Multispectral Qu ; a
Table 3 Estimation of width of a road using digital method
4.6 Discrimination of Asphalt and Concrete Road
Concrete and asphalt have different reflectance characteristics. In ideal condition, it is possible to distinguish one from
another using remote sensing data. But, if these materials are used in roads, due to wearing of vehicle tyres, black
coloured rubber particles begin to stick on their surface after the road become operational. As a result, concrete road
surface gives similar kind of reflectance as asphalt.
5 CONCLUSION
Identification of road characteristics depends on its surrounding land cover or background and spatial resolution of the
data. If a road passes through an area having very different spectral signature from its own, such as uniform vegetation
like a paddy field, its edges can be identified clearly. In a background like a water body, edges of a road are also
prominent. Roads are not well identified, if it pass through an urban area with plenty of structural materials. ADEOS
Panchromatic and SPOT Panchromatic data provide more information on a bridge than ADEOS Multispectral or
LANDSAT TM.
International Archives of Photogrammetry and Remote Sensing. Vol. XXXIII, Part B3. Amsterdam 2000. 371