Detection of forest fire scar in peatland using segmentation analysis and pixel-based classification of PlanetScope imagery

Forest fires are the burning of any forest type primarily due to human activities or induced by natural factors and are events where uncontrolled fire spreads rapidly in a forested area. Many image classifications using remotely sensed data techniques use pixel-based classification approaches, including fire scar detection. Segmentation image analysis is the process of partitioning an image into multiple segments and is typically used to locate objects and boundaries. This study was carried out to assess the performance of segmentation and pixel-based classification in forest fire scar detection in the peatland forest by using a 3-metre resolution of PlanetScope imagery. The overall accuracy assessment of five land cover classes in the study area was higher using segmentation image analysis, which was 89%, compared to pixel-based classification, which was 80%. Therefore, this study concluded that segmentation analysis is more accurate in

Actions (login required)

View Item