Landsat image is commonly used in
land use changes (Chen et al., 2003; Robertson and King, 2011; Hassan et al.,
2016). However, a Landsat image with a roughly 30 m spatial resolution often
cannot meet specific requirements in the land-use classification process
especially in complex urban areas (Lu and Weng, 2005).
In order to increase the potential to extract more detailed in land use
changes, high spatial resolution image has been used widely in land use
changes. In the recent decade, researchers have been used of high spatial
resolution images that are better than 5 m spatial resolution such as IKONOS
and Quickbird for changes application (Goetz et al., 2003; Xu et al., 2003; Lu
and Weng, 2009). The issue of difficulty in obtaining cloud-free images from
satellite platform can be overcome by using UAV aerial photography. This is because, most of the UAV is flying at low
altitudes below cloud cover. The use of UAV at low altitude and using small
digital camera is very practical tools for land cover studies and effective for
mapping small study area.
There are many studies that have been compared the object based and pixel
based change detection. For example, Robertson and King (2011), state that object based classification depicted change
more accurately than maximum likelihood classification because object based are
detected more realistic for land use land cover changes with fewer illogical
errors. Furthermore, object based can
capture more meaningful detailed change information which model the actual
geographic entities (Chen et al.,
2011; Castilla and Hay, 2008).
In addition, Zhou et al.,
(2008) revealed that object based provides an effective way to incorporate
spatial information and expert knowledge into the change detection
process. Beside in LULC changes, object based also been proven the effectiveness
in urban land use changes (Herold et al.,
2002), forest death monitoring (De Chant and Kelly, 2009), shrub mapping
(Zahidi et al., 2017) and so on.
Beside classification method, change detection technique also important
in land use land cover changes. Some of the studies such as (Chen et al., 2003; Hassan et al., 2016) have been used post
classification technique to increase the assessment of classification. The post classification approach is comparison analysis of classification images
for different dates (Alagu Raja et al.,
2013). GIS tools were used for the post-classification comparison of the
changes to producing change map between years. The different classified images
were overlaid to produce land use land
cover change map. Peiman (2011) have monitoring
the land cover land use changes by using pre
classification and post classification
change detection for 34 years period.