| Summary: | As time goes by, remote sensing developments have same way with
development of technology especially in sensor and plane. This also extend on
remote sensing application such as vegetation object. The problems for this
method on high spatial resolution image are salt and pepper who appear in result
of classification. The purpose of this research are compare efectivity between
piksel based classification and obyek based classification for composition
vegetation mapping on high resolution image Worldview-2. OBIA has advantages
than pixel based classification because OBIA not just use per-pixel value but
cluster piksel.
Cases study for this research on Hutan Tidar Magelang which are tropical
forest with have vegetation composition heterogen. Field measurement data are
used for re-classification process and accuracy assesment. The location of point
sample are based on the visual classification. Stratified random sampling are
used to determine the location of point sample. Composition vegetation classes
are divide by 10 and 3 more class for uncomposition vegetation class.
The results show that pixel based classification using majority 5x5 kernel
windows give the highest accuracy between another classification. The highest
accuracy is 73,32% from image Worldview-2 are being radiometric corrected
level surface reflectance. But for overal accuracy in every class,object based are
the best between another methods. Piksel based classification better then obyek
based classification from aspect cost time and difficulty,and obyek based are
better then piksel based on overall accuracy. Reviewed from effectivity aspect,
piksel based are more efective then obyek based for vegetation compotition
mapping in Tidar forest.
Keywords: Remote sensing, Objek based classification, Piksel based
classification, Worldview-2, Floristik composition, Effectivity
|
|---|