2.3.1. Object-oriented image segmentation
The TM images were used to extract polygonal assessment
units that contained relatively uniform spectral and
spatial characteristics through object-oriented segmentation for
habitat suitability. The final segmentation parameters were
as follows: color:shape = 8:2, compactness:smoothness = 2:4. The
multi-resolution segmentation algorithm was performed using
Landsat TM4 (near-infrared), TM3 (red), and TM2 (green) images
at five segmentation scales: 5, 10, 20, 30, and 50. After comparing
the results derived from the five segmentation scales, the finerscale
parameter of 20 was used, which was deemed as the most
acceptable.
2.3.1. Object-oriented image segmentationThe TM images were used to extract polygonal assessmentunits that contained relatively uniform spectral andspatial characteristics through object-oriented segmentation forhabitat suitability. The final segmentation parameters wereas follows: color:shape = 8:2, compactness:smoothness = 2:4. Themulti-resolution segmentation algorithm was performed usingLandsat TM4 (near-infrared), TM3 (red), and TM2 (green) imagesat five segmentation scales: 5, 10, 20, 30, and 50. After comparingthe results derived from the five segmentation scales, the finerscaleparameter of 20 was used, which was deemed as the mostacceptable.
การแปล กรุณารอสักครู่..
