Analysis of polarimetric synthetic aperture radar and passive visible light polarimetric imaging data fusion for remote sensing applications.

摘要

The recent launch of spaceborne (TerraSAR-X, RADARSAT-2, ALOS-PALSAR, RISAT) and airborne (SIRC, AIRSAR, UAVSAR, PISAR) polarimetric radar sensors, with capability of imaging through day and night in almost all weather conditions, has made polarimetric synthetic aperture radar (PolSAR) image interpretation and analysis an active area of research. PolSAR image classification is sensitive to object orientation and scattering properties. In recent years, significant work has been done in many areas including agriculture, forestry, oceanography, geology, terrain analysis. Visible light passive polarimetric imaging has also emerged as a powerful tool in remote sensing for enhanced information extraction. The intensity image provides information on materials in the scene while polarization measurements capture surface features, roughness, and shading, often uncorrelated with the intensity image. Advantages of visible light polarimetric imaging include high dynamic range of polarimetric signatures and being comparatively straightforward to build and calibrate. This research is about characterization and analysis of the basic scattering mechanisms for information fusion between PolSAR and passive visible light polarimetric imaging. Relationships between these two modes of imaging are established using laboratory measurements and image simulations using the Digital Image and Remote Sensing Image Generation (DIRSIG) tool. A novel low cost laboratory based S-band (2.4GHz) PolSAR instrument is developed that is capable of capturing 4 channel fully polarimetric SAR image data. Simple radar targets are formed and system calibration is performed in terms of radar cross-section. Experimental measurements are done using combination of the PolSAR instrument with visible light polarimetric imager for scenes capturing basic scattering mechanisms for phenomenology studies. The three major scattering mechanisms studied in this research include single, double and multiple bounce. Single bounce occurs from flat surfaces like lakes, rivers, bare soil, and oceans. Double bounce can be observed from two adjacent surfaces where one horizontal flat surface is near a vertical surface such as buildings and other vertical structures. Randomly oriented scatters in homogeneous media produce a multiple bounce scattering effect which occurs in forest canopies and vegetated areas. Relationships between Pauli color components from PolSAR and Degree of Linear Polarization (DOLP) from passive visible light polarimetric imaging are established using real measurements. Results show higher values of the red channel in Pauli color image (|HH-VV|) correspond to high DOLP from double bounce effect. A novel information fusion technique is applied to combine information from the two modes. In this research, it is demonstrated that the Degree of Linear Polarization (DOLP) from passive visible light polarimetric imaging can be used for separation of the classes in terms of scattering mechanisms from the PolSAR data. The separation of these three classes in terms of the scattering mechanisms has its application in the area of land cover classification and anomaly detection. The fusion of information from these particular two modes of imaging, i.e. PolSAR and passive visible light polarimetric imaging, is a largely unexplored area in remote sensing and the main challenge in this research is to identify areas and scenarios where information fusion between the two modes is advantageous for separation of the classes in terms of scattering mechanisms relative to separation achieved with only PolSAR.