The geometric accuracy of Sentinel-1 is usally derived from level-1 products in Stripmap mode using accurate surveyed corner reflector positions. While the range offset is found to be low (~30 cm) an azimuth offset of about 2 m remains. The study investigates the reason for the geometric offsets by deriving the geometric parameters from L0-data. Therefore the pulse round-trip time is analyzed from transmission to reception of a point target's backscatter including a simple pulse compression in range and azimuth. Finally, the azimuth offset is reduced to about 15 cm which indicates an artificial bias arising from SAR processing.

This paper introduces a boundary detection technique for remotely sensed images using a CNN-based multi-band fusion architecture. Conventional edge detection techniques fail to cope with the improvements in spatial, spectral, and radiometric resolutions of remote sensing images. While current approaches have handled each complexity in a mutually exclusive manner through specific adaptation of boundary detection parameters, there have been limited techniques that are feature-independent and parameter-free. The proposed approach attempts to integrate complementary and redundant information from the various multi-spectral bands of remotely sensed images to provide a composite image which could enhance perception and reinforce common interpretation while facilitating self-learning and customization capabilities for detection and fusion respectively. First, the technique associates a confidence map on the location of region boundaries with each multi-spectral band individually using a Convolutional Neural Network (CNN). Further, a weighted decision-based fusion framework is applied to integrate the contributions of individual confidence maps into one unified boundary detected output. Systematic experiments are conducted on publicly available datasets in order to evaluate the performance of the method and benchmark it against competing baselines.

In today's modern era, with the advent of technology many people rely on GPS systems to find a destination. However, locating a destination within the confines of an indoor environment is still a challenge as GPS signals do not work indoors, and most people still resort to the conventional method of referring to a physical floor plan or directory. This paper sets to present a mobile software application that uses a visual recognition and pedestrian dead reckoning (PDR) system to create an interactive indoor navigation system using augmented reality (AR). In order to provide an intelligent and context-sensitive navigation platform, data from smartphone sensors are utilised to enable the navigation system to work in an indoor environment. These sensors include using the accelerometer, compass and camera - sensors which already exist in today's mobile devices. Using these technologies, a prototype for an Android application called SunMap+ has been developed to showcase the possibilities of indoor navigation using AR in the test environment of Sunway University campus.

This paper deals with the azimuth resolution improvement problems via the coherent combination of the SAR images with nearly zero-overlapped Doppler bandwidth, which corresponding to large along-track baseline. Prior to the combination of the images, the phase bias contributions from the atmospheric propagation delays and baseline measurement errors should be calibrated. The interferogram of the azimuth pre-filtered SAR images are adopted in the classical method, whose performance falls down when the overlapped spectra comes to zero. In this paper, the maximum sharpness optimization is adopted to deal this situation. The proposed method has been tested by processing the real data. (4 pages)

A new method to detect bridge over water in high resolution SAR images is presented. Firstly, 2-dimensional fuzzy feature is utilized to segment water regions in SAR image. Secondly, tree and hill shadow false targets are excluded based on 8-neighbour chain-code tracking edge method. Then the edge points of potential bridge ROIs (regions of interest) obtained by searching the water area are used to line fitting. In order to avoid the effect of stripe jamming, the fitted line segments are extended to gain the suspected bridge ROIs. Finally, the suspected bridge ROIs are segmented further by double parameter CFAR, and the bridges are detected by the principle axis direction and perimeter extracted. Experimental results show the algorithm can detect the bridges over water accurately, especially the bridges with subtle grey gradient variance or disturbed by stripe jamming in high resolution SAR image. (4 pages)

For high resolution digital imagery obtained by linear array imaging systems, principles, working procedures of multi-view matching model, including both GC³ and MVLL model, were introduced. Similarities and differences between GC³ and MVLL model were analyzed. Then, the method and process to integrate GC³ and MVLL model was proposed. Based on ADS40 aerial three-line-array image datasets, experiments were done on a discontinuity feature. The experimental results proved that the integrated model could synthesize more image information to enhance matching reliability, and compared with single multi-view matching model, possessed greater matching superiority.

The precipitation is a very important input in the traditional hydrological model and at the same time caused the uncertainty of the land surface model. Due to the heterogeneity existed in time and space, plus the simplexes of traditional methods and the insufficient of precipitation measurement technology, the observed precipitation data restricts the applications of the distributed model in simulation and forecasting to a certain extent. With the recent development of measurements and network technology, the meteorological forecasting has become a friendly-accessed, real-time system which can obtain an extended coverage of the land resources. However, the credibility, accuracy and practicability of the interpreted precipitation data needs to be further observed. Through the image processing, the resources were interpreted into the precipitation data which were compared with the measured rainfall values. Then based on the behavior of the hydrological-forecast resources, the method of credence test was selected, and a reliability analysis was conducted from the aspects of point and side. This paper proves that the method of obtaining point precipitation through the network access has a comparatively low precision, but the simulation of precipitation with measured data in temporal and spatial distribution is well established, which can be used as a supplement to the measured precipitation. So the hydrological-forecast resources mentioned above can be applied to the traditional hydrological model, which will greatly impact the study of precipitation in the temporal and spatial distribution, and improves the precision of the hydrological-forecasting.

We present the results of application high-orbital (GPS/GLONASS) and low-orbital (Parus/TRANSIT) satellite navigation systems to the study of different-scale irregularities in the ionospheric total electron content (TEC), such as troughs, crests of equatorial anomaly, traveling ionospheric disturbances and so on. We also present the results of the comparison of the global ionospheric maps (GIM) of vertical TEC, which are now widely used in ionospheric research, with the results of low- and high-orbital radio tomographic ionospheric imaging and with the data of UV spectral imaging form GUVI instrument (Global Ultraviolet Imager). The data from low-orbital radiotomography systems in Russia (Moscow-Svalbard) and Alaska (Arctic Village-Cordova) were involved in the comparison as well as the data of the IGS (International GNSS Service) network. The comparisons cover the time interval from 2003 to 2008, which includes both geomagnetically quiet and disturbed periods. We also demonstrate the possibilities of GPS/GLONASS TEC studies in connection with solar flares and artificial ionospheric heating. (5 pages)

Remote sensing technology can obtain the urban land cover information quickly and accurately, and it has been widely used in the urban development. In view of the information extraction present in high resolution remote sensing image and database construction characteristics, in this paper, the Google Earth image is data source, an object-oriented method including image segmentation, feature space optimization and the fuzzy classification rules are proposed to extract the urban land-cover information. The precision of the extraction information is 94.17% and Kappa coefficient is 0.8302. The format of urban land-cover information was changed from raster to vector, then which was transferred to GIS software to construct the database. The results show that it has proved the feasibility and practicability to extract the urban land cover information from the high resolution remote sensing images and construct database in the GIS software. (7 pages)

We study the characteristics of massive complex 3D scene data and its remote interactive rendering through networks. Then a new data organization and transfer strategy is proposed. First of all, the server side analysis and partition the 3D scene data source. For each block, an improved SPIHT algorithm is used to code it into compressed data, which is reorganized to form a bit stream for the transmission by networks. The clients adopt a progressive transmission strategy by comprehensive consideration of the needs of the client users, the networks conditions and the data characteristics. Once any length of forehead of a bit-stream is received by the client, it can be added to buffer for 3D scene reconstruction, greatly reducing the time of waiting for display. Then we adopt a LOD scene rendering algorithm based on view-dependent visible region, which reduces the data volume immensely that needs transmission immediately. Simulations results demonstrate that the system is built by our algorithm ensure primely that clients render the scene accurately in real time and users have a good visual experience with low network bandwidth.