การกำหนดพิกัดของภาพถ่ายซาร์ความละเอียดสูงกับภาพถ่ายดาวเทียมระบบออปติคัลด้วยวิธีการจับคู่ภาพ / หทยา ยงบุญธนภัทร = Registration of high resolution SAR images and optical satellite images by features matching method / Hataya Yongbunthanaphat
SAR (Synthetic aperture radar) is a kind of modern technique used to take high-resolution images of wide areas. It can also be used under cloudy weather condition whether it's day or night. With its ability, it is likely that more people tend to use SAR. However, the images produced by SAR are radiation image, so this may cause a problem in image interpretation. Because of that, SAR needs to be used with other kinds of images received from other techniques such as quick bird images in order to have a better analysis of images. In this research, a processing chain to register SAR high resolution images, TerraSAR-X with optical images by using feature matching method in the principle of Template Matching for registration is proposed. From the research, it is found out that using the template matching technique plays a big part in processing 15 features from TerraSAR-X and QuickBird images which have been feature detected and extracted. The average correlation coefficient which shows geometry accuracy of the Template matching method gained from 12 features out of 15 features which pass evaluation result is 0.639.The highest and lowest correlation coefficient are 0.781 and 0.508 respectively.
