Generated Reference Ground Control Points from SAR Image
Sompong Wisetphanichkij, Kobchai Dejhan, Fusak Cheevasuvit, Somsak Mitatha and Somnuk Yartpramot
Facility of Engineering and Resear5ch Center for communication and Information Technology
King Mongkut's Institute of Technology Ladkraband, Ladkrabang,
Bangkok 10520, Thailand
Tel: 66-2-3269967 , 66-2-3269081, Fax : 66-2-3269086
E-mail : kobchai@telelan.telecom.eng.kmitl.ac.th
Chanchai Pienvijarnpong
National Research Council of Thailand
MOS-I Receiving Ground Station , Ladkrabang, Bangkok 10520, Thailand.
Chatcharin Soonyeekan
Faculty of Engineering, Kasem bundit University
Pattanakarn Road, Klongton , Bangkok 10250, Thailand
Abstract
The precision of geometric correction depends on the number of ground contrl points (GCP). The GCPs are seleced from the accentuated features such as the building and the intersection of roads and etc. Anyway, these kinds of features are difficult to define, if and images ot be covered by fog or cloud. On the other hand, it will be clearly appeared on the SAR images, since SAR image obtained form the reflectance of the microwave wavelength which propagates through even the cloud. The coordinate of GCPs of SAR image are obtained by segmenting techniques, such as, gray level slicing, inverse linear transformation and images exclusion . the SSDA ( sequential similarity detection algorithms ) has been adopted to carry out the relationship of GCPs on SAR and other image. Therefore, the GCPs of SAR will be transformed and referred as GCPs reference for geometric correction of any image.
Introduction
To analyze the imagery problem is obtained by remote sensing device, that covers the sensor errors can be broadly categorized as geometric and radiometric distortion.
-Geometric distortions
the sensor errors in this category; are characterized by a geometric or spatial distortion of the image. The error sources are due to the internal ( sensor -related ) and external errors ( such as platform attitude, scene characteristics ), are shown in fig 1.
-Radiometric distortions
Radiometric distortions are characterized by an incorrect intensity distribution , spatial frequency filtering of the scene data , blemishes in the imagery, banding of the images data, and etc. these distortions are caused by camera or scanner shading effects, detector gain variations, atmospheric and sensor induced filtering, sensor imperfections, sensor detector gain errors, and etc. the digital techniques to reduce the effects for radiometric distortion are presented in previous papers [1-3]. In such problem geometric correction can be subdivided into two parts :1) determination of the warping function which will transform the geometry of the scene to the desired geometric coordinate system ; and 2) processing of the digital image intensity samples, given the warping function to produce image samples on the desired coordinate grid. The later process called " resampling " . for precision of geometric correction depends on the number of ground control point (GCPs). The reference GCPs are selected from the accentuated feature such as the building, the intersection of roads and divergence of river/canal, and etc. these kinds of features are clearly appeared on the SAR image. Therefore, the coordinate of GCP of enhanced SAR image simply obtained by gray level slicing as discussed in next section. Section 3 shows the normalized static correlation measurement technique based on Sequential Similarity Detection Algorithms (SSDA) used for finding the relationship during SAR sub image, where contain the GCP and the other image. The GCPs of SAR will be transformed and referred as GCPs reference for geomet4ric correction for any image. Experiments and results are performed and shown in last section.
(a) Typical sensor internal distortions
(b) Typical sensor external image distortions
Fig.1 Characteristic of geometric image distortions
Fig.2 Characteristic radiometric image distortions
