Ghosh, Jayanta Kumar
MAPPING OF TEA GARDENS FROM SATELLITE IMAGES -A FUZZY
KNOWLEDGE-BASED IMAGE INTERPETATION SYSTEM.
Jayanta Kumar GHOSH, Pramod Narayan GODBOLE and Sanjay Kumar GHOSH
Civil Engineering Department. University of Roorkee, Roorkee. 247 667 INDIA.
gjkumfce@rurkiu.ernet.in
KEY WORDS: Automation, Classification, Fuzzy logic, Image interpretation, Land cover, Mapping.
DERE A
ABSTRACT
RSR SERA MERE
Detection and identification leading to interpretation and mapping of tea gardens from satellite images is the
prerequisite for application of remote sensing technology to monitor and management of tea gardens. This paper
discusses the development of a fuzzy knowledge-based image interpretation system for mapping of tea gardens from
satellite images. It emulates the multi-stage, multi-feature and multi-iteration heuristics of an expert image analyst.
Knowledge acquisitions for the system is achieved through spectral knowledge of land covers, domain knowledge
and expert’s heuristics. The knowledge base and feature attributes of the information classes are expressed by
linguistic variables and fuzzy attributes. The inference mechanism is modeled on the basis of fuzzy logic. The
system provides information of different types of land cover in each stage of its interpretation leading to mapping of
tea gardens after the final stage. The mapping of tea gardens from IRS (Indian Remote Sensing Satellite) LISS
(Linear Imaging Self Scanner) II geocoded images of an area in the district of Cacher in Assam (INDIA) has been
carried out by the system developed. The results obtained from the working of the system in each stage of its
operation as well as from the experimental study show that the developed system provides sufficiently accurate
information in each stage of its interpretation. It has been found that the performance of the system is better than the
minimum classification accuracy required i.e., 85% to justify the operational capability of the system. Thus it can be
concluded that the developed system can be used reliably for mapping of tea gardens from satellite images.
|
|
i
i
ÿ
i
p
I
|
|
|
|
1. INTRODUCTION
Tea is one of the most valuable natural resources of India. It commands a pivotal position in the nation's economy
as it is one of the major forex earner for the country. Naturally there is always demand towards cost-effective
techniques for continuous monitoring, assessment and management of tea gardens.
Remote sensing technology offers numerous advantages over traditional methods of conducting agricultural resource
survey and management (Myers, 1983). However, detection and identification leading to interpretation and mapping
of the tea gardens from satellite images is the prerequisite for application of remote sensing technology to monitor
and management of tea gardens. The mapping of tea gardens from satellite data can be carried out either by
photointerpretation or by quantitative analysis.
Ghosh et. al (1992) interpreted IRS LISS II images visually to delineate tea gardens along with other land covers of
a region in Barak Valley of Assam, India. Pal et al (1993) also applied visual interpretation technique to delineate
and assess the condition of tea gardens from satellite images.
However, analysis by photointerpretation method has some serious drawbacks. One of the major limitations lies in
its inconsistency in output, to be more specific, in the segmentation process. This is due to inherent fuzziness of
expert’s cognition and interpretation process as well as that of the satellite data. Moreover, this method can
assimilate only a limited number of distinct brightness levels and has limited multi-spectral analysis capability.
The advances in computer technology have opened the vista for analysing satellite data using computational
methods resulting in consistent output. It has capability to differentiate the full dynamic range of brightness values
and also to analyse the whole range of mutispectral data. It also provides accurate quantitative measure. But
approaches using crisp mathematical models for interpretation of satellite data can neither provide output
comparable to that as given by an expert image analyst nor can it simulate the complex visual image interpretation
process.
460 International Archives of Photogrammetry and Remote Sensing. Vol. XXXIII, Part B7. Amsterdam 2000.
