Extraction of Average Structure Orientation in Urban Area from POLSAR Measurement

It is quite useful to grasp ground spatial aspects and its temporal change. Especially in highly urbanized area such capability has very important role for not only development planning but also disaster recovery planning. Table 1 shows examples and their usages of urban parameters. POLSAR has operational advantages over optical sensors because of acquisition capability regardless of weather, day/night, smoke, clouds and dust. Moreover satellite-borne POLSAR has such features that it can observe periodically basis but cost-effective way, and quite wide area but quickly with enough spatial resolution. It is expected that the urban characteristics such as density and average size of artificial structures, orientation angles of them and so on, can be estimated from such POLSAR signals as an averaged values over areas of interest. For example, it is reported that an arguments of mutual correlation coefficient of circular polarized backscattering components are sensitive to an orientation of artificial structures in area of concern[5]. Moreover other 2nd-order analyzed values such as Entropy and Angle alpha (H-α)[1] or components decomposition methods have been proposed[2-4] as detailed interpretation of targets characteristics. Relationship between urban structure density and the 2nd 0rder values are discussed[6]. In this paper, we propose estimation method of average orientation angle of structures such as buildings or houses by combining such 2nd order analyzed values.