Angular Error Research Articles

Citrus pose estimation from an RGB image for automated harvesting

Automated fruit harvesting is promising research in the development of agricultural modernization. However, the complex and non-structural orchard environment is extremely challenging. In order to meet the needs of different end-effectors and to improve the success rate of automatic fruit harvesting, it is critical to perform fruit pose estimation before picking operations. In this study, a citrus pose estimation method through a single RGB image is introduced. The rotation of the citrus pose is defined as a vector that passes through the center of the fruit, which is perpendicular to the plane where the fruit navel point is located. Simply speaking, a multi-task learning model named FPENet is proposed to simultaneously locate the fruit navel point and predict the fruit rotation vector. And a hyperparameter is introduced in the loss function to achieve the simultaneous convergence of multiple tasks. In addition, this paper designs a 2D image annotation tool and constructs a citrus pose dataset, which contributes to model training and also the algorithm evaluation. In the experiment, we evaluate and analyze each module of the proposed network structure, and verify its performance on a harvesting robot. The experimental results show that the FPENet achieves an 88.92 AP score on fruit navel point detection, and 11.13° on the average error of the rotation vector. Over 90% of rotation vectors have an angular error of less than 22.5°. The harvesting success rate is 79.79%. This study offers a new idea for fruit pose estimation and provides the possibility and foundation for estimating fruit pose with a 2D image input.

Data-driven denoising of stationary accelerometer signals

Modern navigation solutions are largely dependent on the performances of the standalone inertial sensors, especially at times when no external sources are available. During these outages, the inertial navigation solution is likely to degrade over time due to instrumental noises sources, particularly when using consumer low-cost inertial sensors. Conventionally, model-based estimation algorithms are employed to reduce noise levels and enhance meaningful information, thus improving the navigation solution directly. However, guaranteeing their optimality often proves to be challenging as sensors performance differ in manufacturing quality, process noise modeling, and calibration precision. In the literature, most inertial denoising models are model-based when recently several data-driven approaches were suggested primarily for gyroscope measurements denoising. Data-driven approaches for accelerometer denoising task are more challenging due to the unknown gravity projection on the accelerometer axes. To fill this gap, we propose several learning-based approaches and compare their performances with prominent denoising algorithms, in terms of pure noise removal, followed by stationary coarse alignment procedure. Based on the benchmarking results, obtained in field experiments, we show that the learning-based models outperform traditional signal processing filtering in terms of pure inertial signal reconstruction. Moreover, they are shown to improve angular errors by one order of magnitude, given a navigation-related task.

Polarimetric monocular leaf normal estimation model for plant phenotyping

We develop in this paper an accurate, pixel-level, plant leaf normal estimation model by a single polarization image. Though traditional sensing can generate leaf 3-D surface information, these techniques are limited in application because of sensor cost and acquisition-time-related considerations. To achieve the detailed and monocular estimation capacity, we propose a novel surface polarization reflection model that considers a mixture of diffuse or specular reflections and describes the actual reflection process more accurately than prevailing models. Our model also directly corresponds with the recorded polarization states, allowing for direct implementation, no additional computational cost, and no requirement for prior knowledge. We also propose a new strategy to disambiguate the normal solution associated with polarization-based imaging. In contrast to existing methods, which use auxiliary sensory information for the disambiguation, we derive a coarse normal map directly from our image data using an off-the-shelf convolutional neural network. Consequently, we facilitate instantaneous data acquisition, which is essential when modeling dynamic non-rigid objects. Using the coarse normal map as a constraint and optimizing optical smoothness properties makes our estimated outcome more accurate than state-of-the-art results. Experiments show that our median normal angular error is 5.6°, offering a threefold improvement to current polarimetric methods and equivalent or better than what SfM-MVS methods provide, yet using only a single image. Our leaf orientation map is also more detailed than existing methods while exhibiting robustness to polarization image noise and the guiding depth map quality. Hence, by a single polarization image, we obtain high-quality surface normal data with no additional aid.

A fractional order-based mixture of central Wishart (FMoCW) model for reconstructing white matter fibers from diffusion MRI

This paper introduces an algorithm for reconstructing the brain’s white matter fibers (WMFs). In particular, a fractional order mixture of central Wishart (FMoCW) model is proposed to reconstruct the WMFs from diffusion MRI data. The pseudo super diffusive modality of anomalous diffusion is coupled with the mixture of central Wishart (MoCW) model to derive the proposed model. We have shown results on multiple synthetic simulations, including fibers orientations in 2 and 3 directions per voxel and experiments on real datasets of rat optic chiasm and a healthy human brain. In synthetic simulations, a varying Rician distributed noise levels, σ=0.01−0.09 is also considered. The proposed model can efficiently distinguish multiple fibers even when the angle of separation between fibers is very small. This model outperformed, giving the least angular error when compared to fractional mixture of Gaussian (MoG), MoCW and mixture of non-central Wishart (MoNCW) models.

An iterative method for reference pattern selection in high-resolution electron backscatter diffraction (HR-EBSD).

For high (angular) resolution electron backscatter diffraction (HR-EBSD), the selection of a reference diffraction pattern (EBSP0) significantly affects the precision of the calculated strain and rotation maps. This effect was demonstrated in plastically deformed body-centred cubic and face-centred cubic ductile metals (ferrite and austenite grains in duplex stainless steel) and brittle single-crystal silicon, which showed that the effect is not only limited to measurement magnitude but also spatial distribution. An empirical relationship was then identified between the cross-correlation parameter and angular error, which was used in an iterative algorithm to identify the optimal reference pattern that maximises the precision of HR-EBSD.

Research on Optical Spectrum Processing for Photonic-Assisted Broadband RF Cross- Eye Jamming System

Cross-eye jamming technology has been used in electronic warfare, which attempts to protect a military platform from monopulse tracking radars. The cross-eye jamming technology has the ability of inducing angular errors to the monopulse tracking radars by transmitting two jamming signals with equal amplitudes and opposite phases. At present, high operation frequency and broadband cross-eye jamming system has been rarely demonstrated. Therefore, the present cross-eye jamming systems are hard to jam frequency-agile radar or multi-band radar, whose carrier frequency covers a large spectral range. In this paper, a photonic-assisted broadband radio frequency (RF) cross-eye jamming system is proposed and experimentally demonstrated. To achieve effective jamming effect, the intercepted radar signal is modulated to optical carriers and the phase shift is realized by optical spectrum processing. The relationship between system parameter tolerances and jamming effects are also simulated. Furthermore, the RF transfer function of X-band, Ku-band and K-band has been obtained in the experiments. The experimental results show that the amplitude and phase mismatch are below 2.15 degrees and 0.4 dB, respectively. Calculated cross-eye gain is 39dB.

Methods for Comprehensive Calibration of a Low-Frequency Angular Acceleration Rotary Table

The total harmonic distortion (THD) index and its calculation methods are presented to calibrate the sinusoidal motion of the low-frequency angular acceleration rotary table (LFAART) and make up the incomprehensive evaluation based on the angular acceleration amplitude and frequency error indexes. The THD is calculated from two measurement schemes: a unique scheme combining the optical shaft encoder and the laser triangulation sensor and a regular scheme using the fiber optical gyroscope (FOG). An improved reversing moments recognition method is presented to upgrade the accuracy of solving the angular motion amplitude based on optical shaft encoder output. The field experiment shows that the difference in the THD values achieved using the combining scheme and FOG is within 0.11% when the signal-to-noise ratio of the FOG signal is higher than 7.7 dB, indicating the accuracy of the proposed methods and the feasibility of taking THD as the index.

Dynamic surface event-triggered control for unmanned sailboat path following based on adaptive tanh line-of-sight

In this article, a dynamic surface control algorithm based on the adaptive tanh line-of-sight with speed adjustment is proposed to solve the problem of path-following control of unmanned sailboat with event-triggered input, considering the dynamic model uncertainty and unknown external disturbance. First, the adaptive tanh line-of-sight method is used to improve the following accuracy and robustness, then an extended state observer is used to estimate the external disturbance and the uncertainty of the internal model, the heading error and the yaw angular velocity error are considered to design the recursive sliding mode dynamic surface controller, at the same time, the speed of the unmanned sailboat is constrained to improve its stability. In addition, to reduce the action frequency of the steering gear, the relative threshold event-triggered control is applied to the controller. According to the Lyapunov stability analysis, it is proved that the proposed control scheme can ensure that all errors in the closed-loop system are semi-globally consistent and ultimately bounded, while the Zeno behavior can be avoided. Finally, the effectiveness of the proposed scheme is verified by simulation and comparative experiments.

Adaptive sensorless control for interior permanent magnet synchronious motor based on sliding mode approach

This paper proposes an adaptive sensorless control based on sliding mode approach for Interior Permanent Magnet Synchronous Motor (IPMSM) using the method of the angular position estimation error extraction. The proposed strategy combines a novel Adaptive Super-Twisting Controller (ASTWC) with a novel Adaptive Observer High-Order Sliding Mode (AOHOSM), where: (i) The control and observer gains are parameterized in terms of a single parameter simplifying its implementation and reducing the tuning time. (ii) By using an auxiliary system, which does not depend on machine parameters, an AOHOSM is designed for estimating the angular position, speed and acceleration in a wide speed range of the IPMSM, (iii) and these are used in an ASTWC to track a desired reference of speed and direct-axis current. (iv) Sufficient conditions are given to ensure the stability of the closed-loop system using a Lyapunov approach. Furthermore, the proposed strategy is validated throughout experimental set-up in order to show its effectiveness. Finally, a comparative study between the proposed strategy and other strategies published in the literature is addressed.

Gaze Estimation via Strip Pooling and Multi-Criss-Cross Attention Networks

Deep learning techniques for gaze estimation usually determine gaze direction directly from images of the face. These algorithms achieve good performance because face images contain more feature information than eye images. However, these image classes contain a substantial amount of redundant information that may interfere with gaze prediction and may represent a bottleneck for performance improvement. To address these issues, we model long-distance dependencies between the eyes via Strip Pooling and Multi-Criss-Cross Attention Networks (SPMCCA-Net), which consist of two newly designed network modules. One module is represented by a feature enhancement bottleneck block based on fringe pooling. By incorporating strip pooling, this residual module not only enlarges its receptive fields to capture long-distance dependence between the eyes but also increases weights on important features and reduces the interference of redundant information unrelated to gaze. The other module is a multi-criss-cross attention network. This module exploits a cross-attention mechanism to further enhance long-range dependence between the eyes by incorporating the distribution of eye-gaze features and providing more gaze cues for improving estimation accuracy. Network training relies on the multi-loss function, combined with smooth L1 loss and cross entropy loss. This approach speeds up training convergence while increasing gaze estimation precision. Extensive experiments demonstrate that SPMCCA-Net outperforms several state-of-the-art methods, achieving mean angular error values of 10.13° on the Gaze360 dataset and 6.61° on the RT-gene dataset.

Accuracy and safety of robotic navigation-assisted distraction osteogenesis for hemifacial microsomia.

This study aimed to verify the accuracy and safety of distraction osteogenesis for hemifacial microsomia assisted by a robotic navigation system based on artificial intelligence. The small sample early-phase single-arm clinical study, available at http://www.chictr.org.cn/index.aspx, included children aged three years and older diagnosed with unilateral hemifacial microsomia (Pruzansky-Kaban type II). A preoperative design was performed, and an intelligent robotic navigation system assisted in the intraoperative osteotomy. The primary outcome was the accuracy of distraction osteogenesis, including the positional and angular errors of the osteotomy plane and the distractor, by comparing the preoperative design plan with the actual images one week postoperatively. Perioperative indicators, pain scales, satisfaction scales, and complications at one week were also analyzed. Four cases (mean 6.5 years, 3 type IIa and 1 type IIb deformity) were included. According to the craniofacial images one week after surgery, the osteotomy plane positional error was 1.77 ± 0.12 mm, and the angular error was 8.94 ± 4.13°. The positional error of the distractor was 3.67 ± 0.23 mm, and the angular error was 8.13 ± 2.73°. Postoperative patient satisfaction was high, and no adverse events occurred. The robotic navigation-assisted distraction osteogenesis in hemifacial microsomia is safe, and the operational precision meets clinical requirements. Its clinical application potential is to be further explored and validated.

Three Dimensional Shape Reconstruction via Polarization Imaging and Deep Learning

Deep-learning-based polarization 3D imaging techniques, which train networks in a data-driven manner, are capable of estimating a target’s surface normal distribution under passive lighting conditions. However, existing methods have limitations in restoring target texture details and accurately estimating surface normals. Information loss can occur in the fine-textured areas of the target during the reconstruction process, which can result in inaccurate normal estimation and reduce the overall reconstruction accuracy. The proposed method enables extraction of more comprehensive information, mitigates the loss of texture information during object reconstruction, enhances the accuracy of surface normal estimation, and facilitates more comprehensive and precise reconstruction of objects. The proposed networks optimize the polarization representation input by utilizing the Stokes-vector-based parameter, in addition to separated specular and diffuse reflection components. This approach reduces the impact of background noise, extracts more relevant polarization features of the target, and provides more accurate cues for restoration of surface normals. Experiments are performed using both the DeepSfP dataset and newly collected data. The results show that the proposed model can provide more accurate surface normal estimates. Compared to the UNet architecture-based method, the mean angular error is reduced by 19%, calculation time is reduced by 62%, and the model size is reduced by 11%.

Single and multiple illuminant estimation using convex functions

The lighting situation in which a picture was taken has an impact on its color. Illuminant estimation is crucial in computer vision because the colors of objects vary as illumination changes. For this reason, numerous methods for estimating the illuminant have been suggested. In this paper, we suggest a novel statistic-based method for estimating single and multiple illuminants using convex functions. In this respect, convex functions are used in the two subsequent steps of normalization and weight creation. After using weighted K-means to segment the picture, each segment’s associated illuminations are determined. The illumination map for the input image is estimated as a final stage. In this study, we also analyze the effect of convexity on color constancy algorithms and present proofs for the convexity of some statistic-based algorithms. Four different single and multi-illuminant datasets have been used to evaluate the proposed algorithm in terms of two evaluation metrics; recovery and reproduction angular error. We believe that the proposed method could be considered one of the statistical state-of-the-art algorithms. In addition, it has competitive results when compared to most learning-based and deep-learning methods. Further advantages of the proposed algorithm include its simplicity of implementation and low execution time.

Quantum digital signature with unidimensional continuous-variable against the measurement angular error.

The continuous-variable quantum digital signature (CV-QDS) scheme relies on the components of quantum key generation protocol (KGP) to negotiate classical signature, which is more compatible with optical fibers. Nevertheless, the measurement angular error of heterodyne detection or homodyne detection will cause security issues when performing KGP in the distribution stage. For that, we propose to utilize unidimensional modulation in KGP components, which only requires to modulate single quadrature and without the process of basis choice. Numerical simulation results show that the security under collective attack, repudiation attack and forgery attack can be guaranteed. We expect that the unidimensional modulation of KGP components could further simplify the implementation of CV-QDS and circumvent the security issues caused by the measurement angular error.

Sun-pointing safe control for atmospheric environment monitoring satellite using magnetic actuation

The design of sun-pointing safe controller is an important part of spacecraft attitude determination and control system. In this work, a new spin stabilization algorithm using only magnetic torquers is proposed. This algorithm has been developed for the atmospheric environment monitoring satellite, which completes the missions of fixed axis rotation and tracking the sun orientation. Firstly, a Lyapunov function, combining the angular momentum error and angular velocity, is formulated and a new sun-pointing attitude control scheme is derived. Then, it is proved that the magnetic control law is asymptotically stable in the absence of disturbance torques. The proposed method is compared with a state-of-the-art fault-tolerant magnetic spin stabilizing controller design method by numerical simulations. Simulation results show that the proposed method has better performance than the state-of-the-art technique. Finally, the applicability and effectiveness of the proposed control method are demonstrated by the hardware-in-the-loop simulation of the satellite.

The inclusion and role of micro mechanical residual stress on deformation of stainless steel type 316L at grain level

Validating crystal plasticity models requires careful consideration of all aspects. The initial conditions are important when comparing a model and experiment, as the initial residual stresses within the material can be significant but are often overlooked due to experimental limitations. Therefore, their inclusion has the potential to improve model predictions. This work explores the efficacy of using type-III residual elastic stresses, measured using high resolution electron backscatter diffraction (HR-EBSD),as pre-test stress distributions in 316L stainless steel. Two methods of processing the stresses collected (direct and using a least squares solver) were used. A existing method was used to incorporate the experimental measurements as initial residual stresses; The model results were compared with each other, a model containing no residual stress and the experiment. The least squares method helped remove extreme stresses calculated by the cross-correlation method at points with high mean angular error. The modelled stress distributions from both the direct and least squares model did not fully match the experimentally observed stresses but similarities were seen within some grains.After loading, little difference was seen between the models, with and without the inclusion of residual stress, implying a negligible effect on the deformation response of the material. However, the predicted stress distribution did not match with the experimentally measured stress distribution after deformation, suggesting further physical effects must be accounted for in crystal plasticity models.

Experiments to identify the optimal sound to use in a new sound ball to improve recruitment, retention, health, and wellness for blind and visually impaired tennis players

Within the sport of Blind and Visually Impaired (BVI) Tennis, the choice of sound is important in locating the ball. We conducted two experiments to choose a sound that will improve the localizability of the ball, in response to a request for new ball development from the International Blind Tennis Association (IBTA). We screened sounds (freesounds.com) for characteristics that the brain best exploits for sound source localization (Risoud et al, 2018). Sample sounds (23) were tested on an outdoor BVI court in a public park using five Bluetooth speakers, and then replicated in an indoor setting; the environments were otherwise naturalistic and unaltered. Blindfolded-sighted participants (n=29) pointed to where they believed sounds originated, by moving an arrow attached to a large protractor. Degree angles were recorded and converted to absolute degree angle error. The standard BVI tennis rattle ball sound resulted in 9.56 degrees of average angular error at a 30-foot distance. After eliminating sounds that 2 or more people either could not hear in either soundscape or that people had degree angle errors over 15 degrees, we discovered a superior localizable sound that resulted in only 4.00 degrees of average angular error at a 30-foot distance.

Three-axis high-accuracy spacecraft attitude estimation via sequential extended Kalman filtering of single-axis magnetometer measurements

Magnetometer is a highly advantageous sensor for determining a spacecraft’s attitude. This article provides a solution to the problem of spacecraft attitude estimation using magnetometer measurements only. To ensure full observability of spacecraft attitude states, it is necessary to use at least two types of sensors. Consequently, utilizing a single sensor, such as the magnetometer, poses a significant challenge for any attitude estimation algorithm, including the extended Kalman filter (EKF). Moreover, implementing the EKF algorithm, or any other attitude estimation algorithm, is computationally intensive. To address these issues, an algorithm has been developed that estimates spacecraft attitude angles and attitude rates using a sequential extended Kalman filter (SEKF). This algorithm offers numerous benefits over those found in the literature such as high accuracy, low computational resource requirements, the ability to converge even with large initial attitude and angular velocity estimation errors, and the ability to function even if two of the three measurement channels of the magnetometer are not functioning. With these benefits, the developed SEKF algorithm is capable of operating in all spacecraft operational modes, delivering accurate performance and computation time. In spite of measurements with large noise values, the high accuracy achieved by the SEKF algorithm enables the magnetometer to serve as the sole source of attitude information, even if one or two magnetometer measurement channels are not functioning.

Adaptive neural network independent joint-based control for an ODE-PDE rigid-flexible manipulator with multiple constraints

This paper deals with the adaptive neural network control of a nonlinear rigid-flexible manipulator subject to external disturbances and multiple constraints. The system dynamic is modeled by ordinary and partial differential equation (ODE and PDE) based on Hamilton’s Principle. An adaptive neural network sliding mode control strategy is designed to estimate the unknown disturbances and uncertain parameter. The sliding surface consisting of angular error, angular velocity error, and end-point vibration information is proposed based on the energy analysis of the flexible beam. The tangent barrier Lyapunov function is used to ensure the angle errors and end-point deflection within the restrictions. It is shown that the joint motion and vibration of the flexible beam can be adjusted only with an independent joint control input, and no boundary force input is required for vibration suppression. The stability is obtained by semi-group theory and LaSalle’s Invariance Principle. Simulation results demonstrate the effectiveness of the control strategy.

Femur reconstruction in 3D ultrasound for orthopedic surgery planning.

Derotation varisation osteotomy of the proximal femur in pediatric patients usually relies on 2-dimensional X-ray imaging, as CT and MRI still are disadvantageous when applied in small children either due to a high radiation exposure or the need of anesthesia. This work presents a radiation-free non-invasive tool to 3D-reconstruct the femur surface and measure relevant angles for orthopedic diagnosis and surgery planning from 3D ultrasound scans instead. Multiple tracked ultrasound recordings are segmented, registered and reconstructed to a 3D femur model allowing for manual measurements of caput-collum-diaphyseal (CCD) and femoral anteversion (FA) angles. Novel contributions include the design of a dedicated phantom model to mimic the application ex vivo, an iterative registration scheme to overcome movements of a relative tracker only attached to the skin, and a technique to obtain the angle measurements. We obtained sub-millimetric surface reconstruction accuracy from 3D ultrasound on a custom 3D-printed phantom model. On a pre-clinical pediatric patient cohort, angular measurement errors were [Formula: see text] and eventually [Formula: see text] for CCD and FA angles, respectively, both within the clinically acceptable range. To obtain these results, multiple refinements of the acquisition protocol were necessary, ultimately reaching success rates of up to 67% for achieving sufficient surface coverage and femur reconstructions that allow for geometric measurements. Given sufficient surface coverage of the femur, clinically acceptable characterization of femoral anatomy is feasible from non-invasive 3D ultrasound. The acquisition protocol requires leg repositioning, which can be overcome using the presented algorithm. In the future, improvements of the image processing pipeline and more extensive surface reconstruction error assessments could enable more personalized orthopedic surgery planning using cutting templates.