A Decoupled Calibration Method Based on the Multi-Output Support Vector Regression Algorithm for Three-Dimensional Electric-Field Sensors

Aiming at the problem that the measured accuracy of the electric field intensity which is affected by the coupling interference by sensor output signal from the component of a three dimensional electric field, the causes of the coupling error was analyzed, and a decoupled calibration method based on support vector regression algorithm for three-dimensional electric field sensor is proposed. The solution of the decoupled calibration matrix was regarded as a multi-objective optimization process, and the optimal decoupling calibration matrix was obtained by the ν-SVR algorithm. The complex inverse calculation of the matrix was avoided, and the calculation error was reduced. A rotary calibration device was designed to accurately measure the angle between the induction electrode of the sensor and the electric-field vector, and an accurate calculation model of the theoretical electric field was established. The experimental results showed that the error between the calculated and theoretical values of the electric-field components obtained by the proposed method were smaller than those obtained by the traditional inverse matrix calibration method, the accuracy of the calibration was improved, the rationality of the calibration method was proven, and the accuracy of the three-dimensional electric-field intensity measurements was further improved.

A Digital-Simulation Model for a Full-Polarized Microwave Radiometer System and Its Calibration

A digital-correlation full-polarized microwave radiometer is an important passive remote sensor, as it can obtain the amplitude and phase information of an electromagnetic wave at the same time. It is widely used in the measurement of sea surface wind speed and direction. Its configuration is complicated, so the error analysis of the instrument is often difficult. This paper presents a full-polarized radiometer system model that can be used to analyze various errors, which include input signal models and a full-polarized radiometer (receiver) model. The input signal models are generated by WGN (white Gaussian noise), and the full-polarized radiometer model consists of an RF front-end model and digital back-end model. The calibration matrix is obtained by solving the overdetermined equations, and the output voltage is converted into Stokes brightness temperature through the calibration matrix. Then, we use the four Stokes parameters to analyze the sensitivity, linearity, and calibration residuals, from which the simulation model is validated. Finally, two examples of error analysis, including gain imbalance and quantization error, are given through a simulation model. In general, the simulation model proposed in this paper has good accuracy and can play an important role in the error analysis and pre-development of the fully polarized radiometer.

Assessment of computer vision methods for motion tracking of planar mechanisms

One of the main challenges on the use of planar mechanisms is to verify and monitor that the trajectories described by the mechanism correspond to those originally required. However, very few research studies have focused on tracking and monitoring the motion of target points located on the mechanisms during operation conditions. In this paper, a comparative study to evaluate the performance of several computer vision methods (CVMs) when used in motion tracking of planar mechanisms is presented. The aim is to compare and identify the best CVM, in terms of precision, speed, low cost, and computational performance, to track the movement of planar mechanisms. For this purpose, a case study corresponding to a planar four-bar mechanism is selected and analysed. The results show that the vision methods based on the homogeneous and non-homogeneous solution of the camera calibration matrix are a technological alternative for monitoring motion trajectories of planar mechanisms.

Prior Compensation Algorithm for Cerenkov Luminescence Tomography From Single-View Measurements

Cerenkov luminescence tomography (CLT) has attracted much attention because of the wide clinically-used probes and three-dimensional (3D) quantification ability. However, due to the serious morbidity of 3D optical imaging, the reconstructed images of CLT are not appreciable, especially when single-view measurements are used. Single-view CLT improves the efficiency of data acquisition. It is much consistent with the actual imaging environment of using commercial imaging system, but bringing the problem that the reconstructed results will be closer to the animal surface on the side where the single-view image is collected. To avoid this problem to the greatest extent possible, we proposed a prior compensation algorithm for CLT reconstruction based on depth calibration strategy. This method takes full account of the fact that the attenuation of light in the tissue will depend heavily on the depth of the light source as well as the distance between the light source and the detection plane. Based on this consideration, a depth calibration matrix was designed to calibrate the attenuation between the surface light flux and the density of the internal light source. The feature of the algorithm was that the depth calibration matrix directly acts on the system matrix of CLT reconstruction, rather than modifying the regularization penalty items. The validity and effectiveness of the proposed algorithm were evaluated with a numerical simulation and a mouse-based experiment, whose results illustrated that it located the radiation sources accurately by using single-view measurements.

Tidal Calibration of Multicomponent Borehole Strainmeters and Validation of the Method Using Surface Waves

We conducted in-situ calibration of fifteen multicomponent borehole strainmeters deployed in and around the expected focal zones of the Nankai megathrust earthquake. The in-situ calibration method compares tidal strain observed by the borehole strainmeters with predicted tidal strains from the solid Earth’s tide and oceanic tidal loading. Then we obtained a calibration matrix to transfer observed strain data to the regional strain field. We estimated the oceanic tidal loading accurately using a Green’s function, which takes the depth of deployment into consideration. We calculated four sets of calibration matrices using combinations of any three of a group of four gauges as well as a calibration matrix using all four gauges. The estimated calibration matrix was validated by comparing observed seismic strain waves after applying the calibration matrix with theoretical seismic strain waves excited by the 2010 Chile earthquake (Mw 8.8). The in-situ calibration was found to be appropriate for all eleven Ishii-type borehole strainmeters and for one of the four Gladwin Tensor Strainmeters (GTSMs). It was also effective with respect to two shear strains for two of the other three GTSMs.

Effect of Sensor Failure on Dynamometry Calibration

Dynamometers are used to measure integrated fluid dynamic loads such as thrust, torque or side forces. To resolve all of three force and three moment components, multiple embedded force gages are often used. Due to arrangement, static loads, and redundancy, the number of sensor channels can exceed the six degrees of freedom needed to resolve the generalized rigid body forces. This paper considers modeling of the force gages as simple springs to develop an elastic model of the dynamometer. The method was applied to a dynamometer consisting of six three-component force gages arranged in an axisymmetric ring. A calibration matrix based on the elastic model with individual force gage sensitivities was shown to match a full calibration matrix where properly summed force gage voltages were obtained under global load application. The elastic model was then extended to consider calibration matrices where sensors were assumed to fail. In this scenario, several virtual loads were applied to the dynamometer and the calibration matrix was obtained by minimizing the least square error. It was found that nearly half of the sensors could be lost and still a virtual calibration could be applied to the measurements. Extending the least square idea, an actual in-situ calibration matrix was formed by striking the dynamometer with a diverse set of instrumented hammer strikes. This calibration matrix also agreed with the other calibrations at frequencies below where system dynamics become important.

Isotropy Analysis of Parallel Six-Axis Accelerometer on Circular Hyperboloids

This paper introduced the isotropic accelerometer using the circular hyperboloids method, which based on modified Gough-Stewart platform (GSP). By the static model of the accelerometer, the isotropy is defined on the acceleration matrix. On the basis of the isotropy condition, the relationship between isotropy index and geometric parameters of circular hyperboloids was investigated. Calculating the isotropy index by the optimization tool, this paper verified that it is feasible to achieve isotropy for the accelerometer. Then taking mass into account, a case is presented to optimize the parameters to construct isotropic accelerometer on circular hyperboloids. According to the 3D model of isotropic accelerometer, the static characteristic simulation was carried out by the finite element method. Based on the simulation experimental results, the calibration matrix was deduced, and the experimental isotropy index was obtained. Comparing the theoretical and experimental isotropy index, the method of circular hyperboloids was proved to be reliable and valid to construct isotropic accelerometer.

Camera Calibration and Fundamental Matrix Estimation with RANSAC

The main aim of this investigation is to study the camera and scene geometry. We will extimate the camera projection matrix also known as calibration matrix, which maps 3D world coordinates to image coordinates, as well as the fundamental matrix, which relates points in one scene to epipolar lines in another. The camera projection matrix and the fundamental matrix can each be estimated using point correspondences. To estimate the projection matrix (camera calibration), the input is corresponding 2d and 3d points. To estimate the fundamental matrix, the input is corresponding 2d points across the two images. We started out by estimating the projection matrix and the fundamental matrix for a scene with ground truth correspondences. Then we moved on to estimating the fundamental matrix using point correspondences from ORB, which is an alternative to SIFT. We used RANSAC to find the fundamental matrix with the most inliers, we filtered away spurious matches and achieved near perfect point to point matching.