Liao Jingjuan

Use of multifrequency, multipolarization shuttle imaging radar for volcano mapping in the Kunlun Mountains of Western China

Abstract A group of volcanoes northeast of Aksayqin Lake, in the western Kunlun Mountains, China, have been identified on multifrequency, multipolarization spaceborne imaging radar-C/X-band synthetic aperture radar (SIR-C/X-SAR) images. Field observations made on the volcanic morphology and terrain features are described in this paper. Analysis of single-band, single-polarization radar backscatter coefficients (σ°) shows that LHV best discriminates the two types of lava flows (pahoehoe and aa lavas), alluvium, and bedrock. The factors affecting the radar backscatter coefficient also are analyzed. Finally, this paper presents KAr isotopic ages of volcanic samples collected in the field and discusses the volcanism in the area.

A method to segment SAR images based on histogram

The image segmentation is fundamental for many tasks of image processing and classification. Many methods for image segmentation have been developed. The some early segmentation methods are based on histogram, and some methods use local statistics, such as the mean, the standard deviation. Other methods are a combination of edge detection and region growing. It is difficult to determine the interval of the pixel value of a kind of ground object for the methods based on histogram in SAR image segmentation. In this paper, we proposed a method, which is based on histogram, and is similar to methods based on local statistics and region growing to segment the SAR images. Since speckle in SAR images disturbs the process of segmentation, we use an edge-preserving filter to reduce speckle. After the histogram is plotted, it will be noted that it is difficult to determine the interval of the pixel value of a kind of ground object. To determine the different ground object, for a pixel in a window of (2n+1)*(2n+1), the mean of all pixels in the window is M, if absolute value of a pixel value minus M is little T, then let the pixel value equal to M. T is a threshold selected based on the standard deviation and the classes of ground object. After the above process, it will be noted that it is still difficult to determine the interval of the pixel value of a kind of ground object in histogram. We use a data processing method named empirical mode decomposition to process the histogram. The pixel value interval of each ground objects can be determined. Based on the interval of each ground objects, we can segment the image. Application this method to SAR images has shown that the method can effectively segment the SAR images.

Dual‐frequency and quad‐polarization SAR observations in Zhaoqing region, China

Abstract C/X band and HH, HV, VH, VV polarization SAR data of Zhaoqing test site in southern China have been acquired in November, 1993, as part of the GlobeSAR program. These data have been processed using the PCI software (PCI Inc.) and analyzed with the support of the field investigation. This paper presents some scientific results of the SAR applications in the fields of geology, forestry, agriculture, and hydrology. Geologic experiments were conducted to study rock type dielectrics and gold mineralization zones. A forestry exploration experiment aims to investigate the forest type discrimination. The agriculture experiment is conducted to study land cover and crop type classification. The hydrology experiment is mainly for studying the river channel processes and determining the submerged area due to flooding by comparing the GlobeSAR data with CASSAR data. Significant results have been achieved in the four application areas. It is shown that radar remote sensing technology, especially with multi‐ban...