Land Cover Changes Detection Using Region Growing Segmentation

Abstract

This paper try to explore an image segmentation model using the region growing method, the method used to segment landsat (land satelite) images for detecting land cover changes. More than 100 images was used in the works, and they collecting from google earth start from year 2015 to year 2019. The accuracy and effectiveness of developing area segmentation methods was computed using using ROC (receiver ooperational ccharacteristics) and studying changes or deterioration quality of certain areas using Landsat satellite images. By building a model or pattern of changes that occur over a certain period of time, and ultimately we will know how new conditions will occur in the near future. on a foundation based on the pattern as well as what is expected to occur in the future (disaster forcasting) can be seen the information. Finally, from the experiment more white pixel shown more land cover changes happened.