Robust Position Estimation Research Articles

Real-time localization with probabilistic maps and unscented kalman filtering: a dynamic sensor fusion approach

This paper presents a robust position estimation technique for autonomous vehicles navigating urban and highway environments. It employs a probabilistic framework, integrating Monte Carlo simulation and Bayesian filtering via particle filters to represent position estimates beyond Gaussian distributions. To improve accuracy, Unscented Kalman Filtering (UKF) is used to fuse radar and lidar data, facilitating the detection of static, pole-like objects. These objects are matched with landmarks from a detailed reference map generated through 3D lidar scans. The proposed method addresses both lateral and longitudinal localisation challenges, ensuring precise vehicle positioning. Comprehensive simulation tests validate the system's effectiveness, demonstrating reliable performance across various driving conditions.

Robust Attitude Estimation for Low-Dynamic Vehicles Based on MEMS-IMU and External Acceleration Compensation.

Attitude determination based on a micro-electro-mechanical system inertial measurement unit (MEMS-IMU) has attracted extensive attention. The non-gravitational components of the MEMS-IMU have a significant effect on the accuracy of attitude estimation. To improve the attitude estimation of low-dynamic vehicles under uneven soil conditions or vibrations, a robust Kalman filter (RKF) was developed and tested in this paper, where the noise covariance was adaptively changed to compensate for the external acceleration of the vehicle. The state model for MEMS-IMU attitude estimation was initially constructed using a simplified direction cosine matrix. Subsequently, the variance of unmodeled external acceleration was estimated online based on filtering innovations of different window lengths, where the acceleration disturbance was addressed by tradeoffs in time-delay and prescribed computation cost. The effectiveness of the RKF was validated through experiments using a three-axis turntable, an automatic vehicle, and a tractor tillage test. The turntable experiment demonstrated that the angle result of the RKF was 0.051° in terms of root mean square error (RMSE), showing improvements of 65.5% and 29.2% over a conventional KF and MTi-300, respectively. The dynamic attitude estimation of the automatic vehicle showed that the RKF achieves smoother pitch angles than the KF when the vehicle passes over speed bumps at different speeds; the RMSE of pitch was reduced from 0.875° to 0.460° and presented a similar attitude trend to the MTi-300. The tractor tillage test indicated that the RMSE of plough pitch was improved from 0.493° with the KF to 0.259° with the RKF, an enhancement of approximately 47.5%, illustrating the superiority of the RKF in suppressing the external acceleration disturbances of IMU-based attitude estimation.

Autonomous Full 3D Coverage Using an Aerial Vehicle, Performing Localization, Path Planning, and Navigation Towards Indoors Inventorying for the Logistics Domain

Over the last years, a rapid evolution of unmanned aerial vehicle (UAV) usage in various applications has been observed. Their use in indoor environments requires a precise perception of the surrounding area, immediate response to its changes, and, consequently, a robust position estimation. This paper provides an implementation of navigation algorithms for solving the problem of fast, reliable, and low-cost inventorying in the logistics industry. The drone localization is achieved with a particle filter algorithm that uses an array of distance sensors and an inertial measurement unit (IMU) sensor. Navigation is based on a proportional–integral–derivative (PID) position controller that ensures an obstacle-free path within the known 3D map. As for the full 3D coverage, an extraction of the targets and then their final succession towards optimal coverage is performed. Finally, a series of experiments are carried out to examine the robustness of the positioning system using different motion patterns and velocities. At the same time, various ways of traversing the environment are examined by using different configurations of the sensor that is used to perform the area coverage.

FPGA-base object tracking: integrating deep learning and sensor fusion with Kalman filter

This research presents an integrated approach for object detection and tracking in autonomous perception systems, combining deep learning techniques for object detection with sensor fusion and field programmable gate array (FPGA-based) hardware implementation of the Kalman filter. This approach is suitable for applications like autonomous vehicles, robotics, and augmented reality. The study explores the seamless integration of pre-trained deep learning models, sensor data from a depth camera, real-sense D435, and FPGA-based Kalman filtering to achieve robust and accurate 3D position and 2D size estimation of tracked objects while maintaining low latency. The object detection and feature extraction are implemented on a central processing unit (CPU), and the Kalman filter sensor fusion with universal asynchronous receiver transmitter (UART) communication is implemented on a Basys 3 FPGA board that performs 8 times faster compared to the software approach. The experimental result provides the hardware resource utilization of about 29% of look-up tables, 6% of lookup table RAMs (LUTRAM), 15% of Flip-flops, 32% of Block-RAM, 38% of DSP blocks operating at 100 MHz, and 230400 baud rates for the UART. The whole FPGA design executes at 2.1 milliseconds, the Kalman filter executes at 240 microseconds, and the UART at 1.86 milliseconds.

Deep learning reconstruction with uncertainty estimation for [formula omitted] photon interaction in fast scintillator detectors

This article presents a physics-informed deep learning method for the quantitative estimation of the spatial coordinates of gamma interactions within a monolithic scintillator, with a focus on Positron Emission Tomography (PET) imaging. A Density Neural Network approach is designed to estimate the 2-dimensional gamma photon interaction coordinates in a fast lead tungstate (PbWO4) monolithic scintillator detector. We introduce a custom loss function to estimate the inherent uncertainties associated with the reconstruction process and to incorporate the physical constraints of the detector.This unique combination allows for more robust and reliable position estimations and the obtained results demonstrate the effectiveness of the proposed approach and highlights the significant benefits of the uncertainties estimation. We discuss its potential impact on improving PET imaging quality and show how the results can be used to improve the exploitation of the model, to bring benefits to the application and how to evaluate the validity of the given prediction and the associated uncertainties. Importantly, our proposed methodology extends beyond this specific use case, as it can be generalized to other applications beyond PET imaging.

Attitude Determination of Photovoltaic Device by Means of Differential Absorption Imaging

Future wireless power transmission will cover power levels up to kilowatts or more and transmission distances up to the scale of kilometers. With its narrow beam divergence angle, optical wireless power transmission (OWPT) is a promising candidate for such system implementations. In the operation of OWPT, it is necessary to estimate the position, direction (azimuth, elevation), and attitude of the target photovoltaic device before the power supply. The authors have proposed the detection of targets using differential absorption imaging and positioning with a combination of stereo imagery. In the positioning by stereo imagery, a condition regarding the consistency of the left and right images can be defined. This corresponds to the certain value of the exposure time of the image sensor, and this depends on the target’s attitude angle. In this paper, we discuss target attitude estimation using this minimum exposure time at which the integrity measure converges. A physical model was derived under general conditions of target position and experimental configuration. Target attitudes were estimated within an error range of 10 to 15 degrees in approximately 60 degrees range. On the other hand, there is an attitude estimation method based on the apparent size of the target. When using this method to estimate the attitude angle, errors are significantly large for specular and diffuse mixed targets like the PV. The method proposed in this paper is a robust attitude estimation method for the photovoltaic device in OWPT.

A Robust and High-Precision Three-Step Positioning Method for an Airborne SAR Platform

When airborne synthetic aperture radar (SAR) encounters long-time global navigation satellite system (GNSS) denial, the system cannot eliminate inertial navigation system (INS) accumulated drift. Platform positioning technology based on SAR image-matching is one of the important auxiliary navigation methods. This paper proposes a three-step positioning method for an airborne SAR platform, which can achieve the robust and high-precision estimation of platform position and velocity. Firstly, the motion model of the airborne SAR platform is established and a nonlinear overdetermined equation set of SAR Range-Doppler based on the ground-control points set obtained by SAR image-matching is constructed. Then, to overcome the ill-conditioned problem generated by the singular Jacobian matrix when solving the equations directly, a three-step robust and high-precision estimation of platform position and velocity is achieved through singular value decomposition and equation decoupling. Furthermore, the error transfer model of systematic and random platform positioning errors is derived. Finally, a set of semi-physical simulation experiments of airborne SAR is conducted to verify the effectiveness of the positioning method and the accuracy of the error model presented in this paper.

Wideband TDoA Positioning Exploiting RSS-Based Clustering.

The accuracy of radio-based positioning is heavily influenced by a dense multipath (DM) channel, leading to poor position accuracy. The DM affects both time of flight (ToF) measurements extracted from wideband (WB) signals-specifically, if the bandwidth is below 100 MHz-as well as received signal strength (RSS) measurements, due to the interference of multipath signal components onto the information-bearing line-of-sight (LoS) component. This work proposes an approach for combining these two different measurement technologies, leading to a robust position estimation in the presence of DM. We assume that a large ensemble of densely-spaced devices is to be positioned. We use RSS measurements to determine "clusters" of devices in the vicinity of each other. Joint processing of the WB measurements from all devices in a cluster efficiently suppresses the influence of the DM. We formulate an algorithmic approach for the information fusion of the two technologies and derive the corresponding Cramér-Rao lower bound (CRLB) to gain insight into the performance trade-offs at hand. We evaluate our results by simulations and validate the approach with real-world measurement data. The results show that the clustering approach can halve the root-mean-square error (RMSE) from about 2 m to below 1 m, using WB signal transmissions in the 2.4 GHz ISM band at a bandwidth of about 80 MHz.

Robust Attitude and Heading Estimation under Dynamic Motion and Magnetic Disturbance.

Robust and accurate attitude and heading estimation using Micro-Electromechanical System (MEMS) Inertial Measurement Units (IMU) is the most crucial technique that determines the accuracy of various downstream applications, especially pedestrian dead reckoning (PDR), human motion tracking, and Micro Aerial Vehicles (MAVs). However, the accuracy of the Attitude and Heading Reference System (AHRS) is often compromised by the noisy nature of low-cost MEMS-IMUs, dynamic motion-induced large external acceleration, and ubiquitous magnetic disturbance. To address these challenges, we propose a novel data-driven IMU calibration model that employs Temporal Convolutional Networks (TCNs) to model random errors and disturbance terms, providing denoised sensor data. For sensor fusion, we use an open-loop and decoupled version of the Extended Complementary Filter (ECF) to provide accurate and robust attitude estimation. Our proposed method is systematically evaluated using three public datasets, TUM VI, EuRoC MAV, and OxIOD, with different IMU devices, hardware platforms, motion modes, and environmental conditions; and it outperforms the advanced baseline data-driven methods and complementary filter on two metrics, namely absolute attitude error and absolute yaw error, by more than 23.4% and 23.9%. The generalization experiment results demonstrate the robustness of our model on different devices and using patterns.

A robust attitude estimation algorithm for seabed inverted ultra-short baseline

The ultra-short baseline(USBL) is fixed upside down on the seafloor to position an autonomous underwater vehicle, which is the inverted ultra-short baseline (i-USBL). Three attitude angles between the i-USBL coordinate system and the navigation coordinate system will be produced when the USBL is installed on the seabed. Significant positioning error will occur if the USBL attitude is undefined. Besides, it also affects the accuracy of attitude angles estimation that the outliers produced when using USBL. In this paper, we propose a robust singular value decomposition (RSVD) method for estimating the attitude angles. Specifically, we develop an attitude calibration model based on the i-USBL positioning mode. The USBL attitude angles is solved using singular value decomposition (SVD) according to the least squares method. The new attitude angles are used to calculate the residuals of the coordinates at each measurement point. The residuals are used to calculate the weights of each point by the IGG III weighting function. The recalculation of attitude angles is performed by using up-to-date weights and SVD. The performance of SVD, the conventional Gaussian Newton iterative method (GN), and RSVD are compared in the simulations. The simulations show that the RSVD has high calibration accuracy and robustness.

Non-Cooperative Target Attitude Estimation Method Based on Deep Learning of Ground and Space Access Scene Radar Images

Determining the attitude of a non-cooperative target in space is an important frontier issue in the aerospace field, and has important application value in the fields of malfunctioning satellite state assessment and non-cooperative target detection in space. This paper proposes a non-cooperative target attitude estimation method based on the deep learning of ground and space access (GSA) scene radar images to solve this problem. In GSA scenes, the observed target satellite can be imaged not only by inverse synthetic-aperture radar (ISAR), but also by space-based optical satellites, with space-based optical images providing more accurate attitude estimates for the target. The spatial orientation of the intersection of the orbital planes of the target and observation satellites can be changed by fine tuning the orbit of the observation satellite. The intersection of the orbital planes is controlled to ensure that it is collinear with the position vector of the target satellite when it is accessible to the radar. Thus, a series of GSA scenes are generated. In these GSA scenes, the high-precision attitude values of the target satellite can be estimated from the space-based optical images obtained by the observation satellite. Thus, the corresponding relationship between a series of ISAR images and the attitude estimation of the target at this moment can be obtained. Because the target attitude can be accurately estimated from the GSA scenes obtained by a space-based optical telescope, these attitude estimation values can be used as training datasets of ISAR images, and deep learning training can be performed on ISAR images of GSA scenes. This paper proposes an instantaneous attitude estimation method based on a deep network, which can achieve robust attitude estimation under different signal-to-noise ratio conditions. First, ISAR observation and imaging models were created, and the theoretical projection relationship from the three-dimensional point cloud to the ISAR imaging plane was constructed based on the radar line of sight. Under the premise that the ISAR imaging plane was fixed, the ISAR imaging results, theoretical projection map, and target attitude were in a one-to-one correspondence, which meant that the mapping relationship could be learned using a deep network. Specifically, in order to suppress noise interference, a UNet++ network with strong feature extraction ability was used to learn the mapping relationship between the ISAR imaging results and the theoretical projection map to achieve ISAR image enhancement. The shifted window (swin) transformer was then used to learn the mapping relationship between the enhanced ISAR images and target attitude to achieve instantaneous attitude estimation. Finally, the effectiveness of the proposed method was verified using electromagnetic simulation data, and it was found that the average attitude estimation error of the proposed method was less than 1°.

Robust Attitude Estimation Using Magnetic and Inertial Sensors

Lightweight and low-cost magnetic and inertial sensors are commonly used for attitude estimation in a wide range of applications, from motion tracking to autonomous navigation. However, the inherent and external sensor errors and the accelerations due to the actual motion of the platform highly affect the estimation accuracy. This study proposes a robust attitude estimation algorithm that compensates the sensor errors and the external accelerations at the algorithm level. The attitude estimation filter, which is structured as a Multiplicative Extended Kalman Filter (MEKF), is modified to compensate for both the long-term and short-term measurement uncertainties. Modification is applied mainly by estimating and compensating the bias and external accelerations for long-term uncertainties and tuning the measurement noise covariance matrix for short-term uncertainties. Simulations test the proposed algorithm, and the results are compared with benchmark algorithms.

Robust attitude estimation for an unmanned aerial vehicle using multiple GPS receivers

Estimation of the system state is an important task in the field of control, localization, and navigation. Many sensors are used to extract the attitude information. GPS (Global Positioning System) is widely used to provide position and velocity for ground and aerial vehicles. However, only a few research works have addressed the problem of attitude estimation using multiple GPS receivers. The main contribution of this paper consists on proposing a robust algorithm for attitude estimation of an unmanned aerial vehicle using INS/3GPS integration. First, we propose two alternative methods for attitude determination with three GPS receivers: Direct Computing Method (DCM) and Least Square Method (LSM) based on Singular Value Decomposition (SVD). Second, the attitude given by GPS receivers is fused with the measurement of a three-axis gyroscope using a robust Smooth Variable Structure Filter (SVSF). The optimal geometric configuration of GPS antennas is determined based on a deep performance analysis of the proposed system. Then, robustness analysis is assessed against measurement noise and parameters uncertainties. Simulation results are presented to demonstrate the robustness of the proposed approach. A more realistic evaluation of the proposed solution is obtained using an aerial vehicle type “Helicopter” from the Virtual Robot Experimentation Platform (VREP simulator). The proposed filter is compared with an Extended Kalman Filter (EKF). The obtained results have confirmed how much the SVSF is robust for attitude estimation in presence of the considered disturbances.

LiDAR- and Radar-Based Robust Vehicle Localization with Confidence Estimation of Matching Results

Localization is an important technology for autonomous driving. Map-matching using road surface pattern features gives accurate position estimation and has been used in autonomous driving tests on public roads. To provide highly safe autonomous driving, localization technology that is not affected by the environment is required. In particular, in snowy environments, the features of the road surface pattern may not be used for matching because the road surface is hidden. In such cases, it is necessary to construct a robust system by rejecting the matching results or making up for them with other sensors. On the other hand, millimeter-wave radar-based localization methods are not as accurate as LiDAR-based methods due to their ranging accuracy, but it has successfully achieved autonomous driving in snowy environments. Therefore, this paper proposes a localization method that combines LiDAR and millimeter-wave radar. We constructed a system that emphasizes LiDAR-based matching results during normal conditions when the road surface pattern is visible and emphasizes radar matching results when the road surface is not visible due to snow cover or other factors. This method achieves an accuracy that allows autonomous driving to continue regardless of normal or snowy conditions and more robust position estimation.

INDOOR ATTITUDE ESTIMATION USING EQUIPPED GYROSCOPES AND DEPTH SENSORS

Abstract. Attitude estimation is central to a wide range of applications such as robotics, virtual reality and mobile smart devices. With the development of sensor technologies, these application devices are often equipped with gyroscopes and depth sensors. In this paper, we propose a novel method to fuse gyroscope and depth information for drift-free and robust attitude estimation in structured indoor applications. Our method relies on the depth information and the Manhattan world assumption to estimate the absolute orientation, which is then fused to correct the accumulated error of the gyroscope-determined attitude. We first utilize the mean shift algorithm on the unit sphere to align the surface normals from the depth measurements with the orthogonal planar structures of the Manhattan world. Therefore, the orientation estimates are drift-free and absolute with respect to the Manhattan world. We then fuse the orientation estimates with the gyroscope measurements in an error-state Kalman filter manner to further improve the attitude estimation accuracy and robustness. We validate the performance of our method on public datasets, demonstrating the robustness and accuracy of the method for attitude estimation.

Position and Orientation Control of a Mobile Robot based on Robust Marker Position Estimation using a Sphere Camera

- This paper describes position and orientation control of a mobile robot using a sphere camera. While a sphere camera has a wider field of view than a normal camera, the field of view is distorted. Therefore, image recognition is performed by compensating for this distortion. The control of a mobile robot using this sphere camera enables better localization estimation, distance recognition, and object recognition than with a normal camera. In this paper, we propose a position estimation method using markers with color invariants that are robust to illumination. We use a hybrid marker system that employs a circle marker and an AR marker. We demonstrate the effectiveness of our method by conducting experiments on position control of a two-wheel mobile robot using the sphere camera and marker recognition.

Electromagnetic Non-Contacting Position Estimation of a Centimetre-Scale Robot

This paper develops an electromagnet-based position estimation system for a cm-scale robot with two degrees of freedom. The orientation of an external electromagnet is actively controlled in realtime to maximize the magnetic field magnitude at the robot. This results in a monotonic relationship between the magnetic field magnitude and radial distance leading to a simple and robust position estimation system. The radial distance is then estimated using an asymptotically stable nonlinear observer designed using a linear matrix inequality. The analytical principles of the estimation system are first presented using key technical lemmas and proofs. Experimental results are then presented on the verification of the analytical principles and on the performance of the position estimation system.

An Efficient and Low-Latency Deep Inertial Odometer for Smartphone Positioning

The Global Positioning System (GPS) is not an effective solution for pedestrian indoor navigation using embedded devices because of poor signal penetration and high power requirements. Indoor positioning based purely on commercial-grade inertial measurement units (IMUs) may provide a good alternative but their significant noise and random bias must be eliminated. Although Kalman filters and pedestrian dead reckoning helped reduce the errors using IMUs, these methods cannot prevent the divergence of estimated position error. Deep learning was proposed for accurate position estimation with IMUs. Despite the progress, robust position estimation in mobile embedded devices using deep learning has not been established yet because of memory requirements, latency and inaccurate position estimation especially for stationary pedestrians. In this study, we present an efficient embedded deep learning approach for robust, real-time and accurate pedestrian position estimation using commercial Android smartphone IMUs. We first extended a publicly available deep inertial navigation dataset (OxIOD) with stationary data to enhance the positioning accuracy for both steady state and motion. Next, we trained and tested a deep learning architecture that yields a higher positioning accuracy with 50% lower network latency and 31% lower network size compared with earlier deep positioning networks such as IONet. Our real-time tests of the model in an Android smartphone indicated that the extension for the dataset reduces the position shift when the smartphone is stationary. Since our embedded deep learning solution simultaneously decreases the positioning error, latency and memory requirements, the solution paves the way for numerous practical indoor navigation applications.

Semantic VPS for Smartphone Localization in Challenging Urban Environments.

Accurate smartphone-based outdoor localization systems in deep urban canyons are increasingly needed for various IoT applications. As smart cities have developed, building information modeling (BIM) has become widely available. This article, for the first time, presents a semantic Visual Positioning System (VPS) for accurate and robust position estimation in urban canyons where the global navigation satellite system (GNSS) tends to fail. In the offline stage, a material segmented BIM is used to generate segmented images. In the online stage, an image is taken with a smartphone camera that provides textual information about the surrounding environment. The approach utilizes computer vision algorithms to segment between the different types of material class identified in the smartphone image. A semantic VPS method is then used to match the segmented generated images with the segmented smartphone image. Each generated image contains position information in terms of latitude, longitude, altitude, yaw, pitch, and roll. The candidate with the maximum likelihood is regarded as the precise position of the user. The positioning result achieved an accuracy of 2.0 m among high-rise buildings on a street, 5.5 m in a dense foliage environment, and 15.7 m in an alleyway. This represents an improvement in positioning of 45% compared to the current state-of-the-art method. The estimation of yaw achieved accuracy of 2.3°, an eight-fold improvement compared to the smartphone IMU.

Robust attitude estimation of rotating space debris based on virtual observations of neural network

SummaryHigh precise estimation and prediction of the target's attitude motion are key technologies for capturing and removing rotating space debris. In this article, a neural‐network‐enhanced Kalman filter (NNEKF) is proposed to improve the precision and robustness of attitude estimation algorithm. The main innovation of the NNEKF is to utilize virtual observations of the inertia characteristics to improve the filter's performances. The virtual observations are obtained using a neural network, which is offline trained using simulation data. In order to decrease the number of nodes of the network, the input data are preprocessed using the discrete Fourier transformation method. Moreover, by involving the characteristic frequencies in the input vector, the neural network can extract information from all the past observations, so as to grasp long‐term characteristics of the dynamical system. Therefore, the NNEKF can provide more precise estimation of the target's moment of inertia, and furthermore improve the accuracy and robustness of attitude estimation and prediction. Simulation results indicate that the NNEKF can reduce the estimation errors by 39% compared with the conventional EKF method when using the same measurement data. And the accumulation errors of prediction using estimates of the NNEKF is just as 24% as the conventional EKF.