Land covers classification using object-based image analysis (OBIA) technique in Tumpat, Kelantan

Nur Shahanim Shahadan, (2018) Land covers classification using object-based image analysis (OBIA) technique in Tumpat, Kelantan. [Undergraduate Final Year Project Report] (Submitted)

Full text not available from this repository.


Land Use Land Cover Map using Object Based Image Analysis (OBIA) is widely used by many researchers in ecology, natural resource management and others. This study was conducted to analyze the land cover classification from Landsat-8 OLI using OBIA technique. This study was carried out in Tumpat area with a total area cover is 13,253.60 hectares. The Landsat-8 OLI was pre-processed to minimize image disturbances. Then, the images was segmented and merge using Multiresolution Segmentation (MRS) and land cover was then classified using Nearest Neighbor (NN) classification with three different scale parameter, shape and compactness. Scale segmentation using scale parameters 50, shape 0.2 and compactness 0.8 was selected to discuss in detail as the segmentation at this level produced highest accuracy of 96.32%. Among four land cover classes, vegetation is the major land cover in the district of Tumpat with a total area of 5432.49 ha (45.42%) and the lowest area class was bare or open land with a total area of 55.58 ha (0.46%). This study showed that OBIA techniques for a medium resolution Image Landsat able to analyze the Land Use Land Cover in Tumpat Kelantan area. However, to improve this study, the uses of high-resolution images like IKONOS or Quickbird will improve the classification results in OBIA techniques to produce better results compared to medium resolution images. The accuracy of the result will be enhanced by implement appropriately sampling technique and the best selection of segmentation scale.

Item Type: Undergraduate Final Year Project Report
Faculty: Faculty of Earth Sciences
Depositing User: En. Shahrul Afzan Ibrahim
Date Deposited: 05 May 2019 08:42
Last Modified: 05 May 2019 08:43
Repository Statistic: View Download Statistic

Actions (login required)

View Item View Item