Segmentation of Synthetic Aperture Radar (SAR) Image Using Graphcut

• Main Author: Ismail, Muhammad Harris Fahmi
• Format: Final Year Project
• Language: English
• Published: Universiti Teknologi PETRONAS 2016
• Subjects: TK Electrical engineering. Electronics Nuclear engineering
• Online Access: http://utpedia.utp.edu.my/22553/1/FYP_16299_FINAL%20DISSERTATION.pdf; http://utpedia.utp.edu.my/22553/

SAR imaging technique is popular for remote sensing and monitoring application as it offers the advantage of high resolution imagery and usability under various weather conditions. Segmentation is important imaging processing steps in interpretation and classification of SAR images. The process of segmentation involving two major steps, namely feature extraction and clustering. Improving the quality of the features will ease the burden on the clustering scheme and to solve this, image filtering is introduced in the mothodology to improve the quality of the images. With the variation of segmentation techniques used, final judgement indicating better technique for SAR images application can be made. Where in this, paper, comparison between two different techniques of segmentation are performed. They are Meanshift Segmentation Technique and Graphcut Segmentation Technique. Evaluations of the experiments are done to get to know which technique are better. Finally, it is come to conclusion that Graphcut Segmentation can perform the segmentation task for SAR images in way fastest time, which is better.