Geography Reference
In-Depth Information
Aperture Radar SAR), which has the ability to acquire remotely sensed imagery
under various weather condition during both day and night. Studies (Solberg et al.
1994
; Huang et al.
2007
) using SAR and optical sensor data have confirmed clear
enhancement in classification accuracies contrary to an optical sensor alone.
Xu and Gong (
2007
) evaluated the potential of the Earth Observing-1 (EO-1)
Hyperion hyper-spectral (HS) data with that of the EO-1 Advanced Land Imager
(ALI) multispectral (MS) data for distinguishing various LULC-classes in Fre-
mont, California.
In addition to the progress achieved by the referenced studies, the use of object-
or segment-based classification techniques is another new development in the
environment of remote sensing image classification. This approach has achieved
generally better success with the narrow bands and high spatial resolution data
such as IKONOS, SPOT-5, or QUICKBIRD (Willhauck
2000
). In several of the
followed studies (e.g., Fuller et al.
2002
; Marcal et al.
2005
; Platt and Rapoza
2008
) segment-based classifications were more accurate than conventional pixel-
based classifications.
2.4 Land Use/Land Cover Change Detection Mapping
Change detection analysis is important in monitoring and managing the natural
resources of the Earth. It gives statistical analysis of the occurred spatial distri-
bution of the LULC-changes of interest (Singh
1989
). Some of its applications are:
Monitoring shifting agriculture, estimation of deforestation, estimation of desert-
ification, and other environmental changes (Jingan et al.
2005
). Natural change can
have a wide impact on natural resources. So, in relation to LULC and natural
resource and ecosystem management, there is an important need for timely, per-
manent, and truthful monitoring of changes occurring. But, the problems chal-
lenging the change detection process are: where is the change?, how much?, when
did it occur?, and how great is its impact on the ecosystem? (Lambin and Lin-
derman
2006
). Changes can occur either suddenly or gradually (Bontemps et al.
2008
). Here, the remote sensing techniques take on an increasing importance in
natural resource monitoring programs and in answering the above questions
(Wiens et al.
2009
). In the case of LULC-changes, two kinds of change can be
classified from previously published literature: conversion and modification
(Lambin et al.
2003
). LULC-conversion is the change from one cover category to
another (e.g., the complete replacement of an agricultural parcel by man-made
buildings), while LULC-modification is the modifications of structure or function
without a complete change from one category to another (e.g., changes in pro-
ductivity, biomass, or phenology).
Search WWH ::
Custom Search