Taro Suzuki

Bio: Taro Suzuki is an academic researcher from Chiba Institute of Technology. The author has contributed to research in topics: GNSS applications & Global Positioning System. The author has an hindex of 13, co-authored 66 publications receiving 571 citations. Previous affiliations of Taro Suzuki include Waseda University & Tokyo University of Marine Science and Technology.

Correcting GNSS Multipath Errors Using a 3D Surface Model and Particle Filter

Abstract: Currently, the availability of global navigation satellite systems (GNSSs) is anticipated to improve because of the presence of various positioning satellites. However, because of the serious impact of multipath signals such as reflection and the diffraction signals caused by geographic features in urban environments, such improvements in the availability of satellite positioning do not necessarily also facilitate high precision positioning. The multipath error effect is highly dependent on the shape and geometry of geographic features (such as tall buildings) near a GNSS receiver and therefore cannot be addressed by differential GNSS techniques that attempt to remove most of the errors in GNSS positioning. Various practical and popular signal correlator techniques can also help mitigate multipath errors. However, when an antenna cannot receive a direct signal, these techniques do not provide satisfactory results because they presume that the antenna principally receives both direct and multipath signals. Urban environments include many locations where non-line-of-sight (NLOS) satellite signals are obstructed by buildings, making it difficult to determine and mitigate such errors from the signals alone. At the ION GNSS 2012 conference, we proposed a novel GNSS positioning technique that can be used in multipath environments [1]. The proposed technique attempts to estimate a user’s position, without pseudoranges containing multipath errors, by comparing GNSS signal strengths that have been simulated using 3D surface models of urban canyon environments. The GNSS signal strength is calculated by the multipath propagation based on a ray-tracing technique. Our results showed that the proposed technique worked well in a real urban canyon environment. In this paper, we further improve this technique. Rather than comparing the signal strengths to estimate user position, we directly correct the observed pseudorange containing multipath errors using the simulated signal delay and strength from the 3D surface model. The multipath error in the observed pseudorange depends on a signal correlator design that is implemented in GNSS receivers. Therefore, to correct the multipath errors using the simulated signal delay and strength, it is necessary to modify the signal correlator of GNSS receivers. Consumer GNSS receivers cannot be used for this purpose, so we use a GNSS software receiver to realize the proposed techniques in this paper. A serious barrier to using multipath simulation techniques to improve real-world positioning accuracy stems from the fact that a precise LOS vector and signal obstruction cannot be calculated when the user position is unknown because ray-tracing results are highly dependent on the user’s position. In other words, to simulate the true multipath error at the user’s position, it is necessary to accurately predetermine that position before simulation. We solve this problem by using a particle filter that provides a numerical approximation of the estimated states of a set of weighted random samples, called particles. Differing user position estimates are represented by these particles. We then evaluate the likelihood of each particle by comparing observed and simulated pseudoranges. Our algorithm is executed as follows: (i) User position estimates are created as particles; the state vectors of these particles are composed of random 2D positions, x and y, with the altitude of each particle calculated from the 3D terrain data. The particles are created at initial positions with a specified distribution, and the mean position of the particles is then determined using a position solution obtained by conventional single-point positioning. The particles created in this initial step are distributed over a large region. (ii) The LOS vector can be computed at each particle position using the satellite positions obtained from broadcast GNSS ephemeris data. We determine the signal refraction or diffraction point on the 3D surface using the relationship between the LOS and 3D model to estimate the signal delay and strength. To reduce complexity and computational load, this paper does not assess multipath signals with several reflections and diffractions. The multipath error at each particle location can be simulated using the estimated signal delay and strength, as well as the configuration of the signal correlator. (iii) For each particle, the simulated pseudorange, which includes simulated multipath errors, is compared with the observed real pseudorange. Particles with simulated pseudoranges close or equal to the actual pseudorange are considered to be the closest to the true position based on a probabilistic model of pseudorange matching. (iv) Finally, the weight of each particle is updated based on its likelihood, and all the particles are resampled based on their new weights. Particles that are sufficiently close to the true position survive this resampling step, and the average position of the remaining particles is taken as the final estimated user position. Using this technique, we can estimate the user position even if the observed pseudorange has a large multipath error or the GNSS receiver receives only NLOS (reflected or diffracted multipath) satellite signals. To confirm the effectiveness of the proposed technique, a positioning test was performed in a real-world urban canyon environment. We set up ground control points, which were predetermined high-accuracy positions, to compare the positioning accuracy. These results show that the proposed technique is effective and offers increased positioning accuracy within urban canyon environments that suffer from large reflection and diffraction multipath errors in GNSS signals. [1] Taro Suzuki, Nobuaki Kubo, GNSS Positioning with Multipath Simulation using 3D Surface Model in Urban Canyon, Proc. of ION GNSS 2012, pp.438-447, 2012.

6-DOF localization for a mobile robot using outdoor 3D voxel maps

Abstract: This paper describes outdoor localization for a mobile robot using a laser scanner and a three-dimensional (3D) voxel map that is based on outdoor point clouds. A mobile mapping system (MMS) measures outdoor 3D point clouds easily and precisely. The complete 6D state of a mobile robot is estimated by combining dead reckoning and the 3D voxel map. The 2D position and orientation are extended to 3D by using the 3D voxel map and by assuming that the mobile robot remains in continuous contact with the road surface. Our approach applies a particle filter to correct position errors in the laser measurement model for a 3D point cloud space. Field experiments were performed to evaluate the accuracy of our proposed method. Our results confirmed that it is possible to achieve a localization precision of 0.2 m (RMS) using our proposed method.

High-accuracy GPS and GLONASS positioning by multipath mitigation using omnidirectional infrared camera

Abstract: This paper describes a precision positioning technique that can be applied to vehicles or mobile robots in urban or leafy environments. Currently, the availability of satellite positioning is anticipated to improve because of the presence of various positioning satellites such as GPS of the U.S., GLONASS of Russia and GALILEO of Europe. However, because of the serious impact of multipath on their positioning accuracy in urban or leafy areas, such improvements in the availability of satellite positioning do not necessarily also facilitate high precision positioning. Our proposed technique mitigates GPS and GLONASS multipath by means of an omnidirectional infrared (IR) camera that can eliminate the need for invisible satellites by using IR images. With an IR camera, the sky appears distinctively dark. This facilitates the detection of the borderline between the sky and the surrounding buildings, which are captured in white, because of the difference in the atmospheric transmittance rates between visible light and IR rays. The proposed technique can automatically and robustly mitigate GPS and GLONASS multipath by excluding the invisible satellites. Positioning evaluation was carried out only with visible satellites that have less multipath errors and without using invisible satellites. The evaluation results confirm the effectiveness of the proposed technique and the feasibility of its highly accurate positioning.

GNSS Positioning with Multipath Simulation using 3D Surface Model in Urban Canyon

Abstract: This paper proposes a novel global navigation satellite system (GNSS) positioning technique that can be used by vehicles or pedestrians in urban canyon environments. With the launch of new positioning satellites, the availability of the GNSS is anticipated to improve, although, owing to the serious impact of multipath signals on positioning accuracy within urban canyon environments, the increased availability of satellite positioning will not necessarily correspond to higher precision positioning. Here, we propose a technique based on multipath simulation using a 3D digital surface model (DSM) for obtaining GNSS positioning in urban canyons where there are large reflection and diffraction multipath errors. To calculate a user’s position using multipath simulation, it is necessary to accurately predetermine their position, as the multipath effect is highly dependent on the surrounding obstructions. A particle filter, in which a number of hypotheses of user position are created, is used to solve this problem and allow multipath simulation to estimate the position. In the positioning process, a comparison is made between the signal-to-noise ratio (SNR) estimated by propagation loss within the 3D DSM and the observed SNR. For more correct hypotheses of a particle’s position, these two values of the SNR will converge, as signal propagation will have been estimated more accurately, and such particles are given larger weights. Finally, an accurate user position can be estimated through a resampling process based on these particle weights. In order to confirm the effectiveness of this proposed technique, a static positioning test was performed in an urban canyon. The results of this assessment show that the proposed technique is effective and offers increased positioning accuracy in urban environments in which large reflection and diffraction multipath errors occur in GNSS signals.

A Small UAV for Immediate Hazard Map Generation

Abstract: An unmanned aerial vehicle (UAV) system providing quick response imagery for natural disaster assessment was developed. A prototype UAV and on-board software using a GPS navigation system were developed for this purpose. The flight performance was evaluated by 6DOF simulations and rigorous flight tests. In the proposed system, collected images by an on-board digital camera is transmitted in real-time to a spatial temporal GIS to share the information. The usability of the small UAV system was qualified by a flight experiment in a city-wide emergency evacuation drill.

Overview and Current Status of Remote Sensing Applications Based on Unmanned Aerial Vehicles (UAVs)

Abstract: Remotely Piloted Aircraft (RPA) is presently in continuous development at a rapid pace. Unmanned Aerial Vehicles (UAVs) or more extensively Unmanned Aerial Systems (UAS) are platforms considered under the RPAs paradigm. Simultaneously, the development of sensors and instruments to be installed onboard such platforms is growing exponentially. These two factors together have led to the increasing use of these platforms and sensors for remote sensing applications with new potential. Thus, the overall goal of this paper is to provide a panoramic overview about the current status of remote sensing applications based on unmanned aerial platforms equipped with a set of specific sensors and instruments. First, some examples of typical platforms used in remote sensing are provided. Second, a description of sensors and technologies is explored which are onboard instruments specifically intended to capture data for remote sensing applications. Third, multi-UAVs in collaboration, coordination, and cooperation in remote sensing are considered. Finally, a collection of applications in several areas are proposed, where the combination of unmanned platforms and sensors, together with methods, algorithms, and procedures provide the overview in very different remote sensing applications. This paper presents an overview of different areas, each independent from the others, so that the reader does not need to read the full paper when a specific application is of interest

GNSS Position Integrity in Urban Environments: A Review of Literature

Abstract: Integrity is one criteria to evaluate GNSS performance, which was first introduced in the aviation field. It is a measure of trust which can be placed in the correctness of the information supplied by the total system. In recent years, many GNSS-based applications emerge in the urban environment including liability critical ones, so the concept of integrity attracts more and more attention from urban GNSS users. However, the algorithms developed for the aerospace domain cannot be introduced directly to the GNSS land applications. This is because a high data redundancy exists in the aviation domain and the hypothesis that only one failure occurs at a time is made, which is not the case for the urban users. The main objective of this paper is to provide an overview of the past and current literature discussing the GNSS integrity for urban transport applications so as to point out possible challenges faced by GNSS receivers in such scenario. Key differences between integrity monitoring scheme in aviation domain and urban transport field are addressed. And this paper also points out several open research issues in this field.

Comparing fully convolutional networks, random forest, support vector machine, and patch-based deep convolutional neural networks for object-based wetland mapping using images from small unmanned aircraft system

Abstract: Deep learning networks have shown great success in several computer vision applications, but its implementation in natural land cover mapping in the context of object-based image analysis (OBIA) is...

Recent applications of unmanned aerial imagery in natural resource management

Abstract: Unmanned aerial vehicles have become popular platforms for remote-sensing applications, particularly when spaceborne technology, manned airborne techniques, and in situ methods are not as efficient...