A Refined Technique to Calculate Finite Helical Axes From Rigid Body Trackers

2014 ◽  
Vol 136 (12) ◽  
Author(s):  
Stewart D. McLachlin ◽  
Louis M. Ferreira ◽  
Cynthia E. Dunning
Keyword(s):  
Rigid Body ◽  
Window Size ◽  
Moving Window ◽  
Data Set ◽  
Hinge Joint ◽  
Starting Position ◽  
Window Analysis ◽  
Plane Normal ◽  
Moving Window Analysis ◽  
Direction Cosines

Finite helical axes (FHAs) are a potentially effective tool for joint kinematic analysis. Unfortunately, no straightforward guidelines exist for calculating accurate FHAs using prepackaged six degree-of-freedom (6DOF) rigid body trackers. Thus, this study aimed to: (1) describe a protocol for calculating FHA parameters from 6DOF rigid body trackers using the screw matrix and (2) to maximize the number of accurate FHAs generated from a given data set using a moving window analysis. Four Optotrak® Smart Markers were used as the rigid body trackers, two moving and two fixed, at different distances from the hinge joint of a custom-machined jig. 6DOF pose information was generated from 51 static positions of the jig rotated and fixed in 0.5 deg increments up to 25 deg. Output metrics included the FHA direction cosines, the rotation about the FHA, the translation along the axis, and the intercept of the FHA with the plane normal to the jig's hinge joint. FHA metrics were calculated using the relative tracker rotation from the starting position, and using a moving window analysis to define a minimum acceptable rotational displacement between the moving tracker data points. Data analysis found all FHA rotations calculated from the starting position were within 0.15 deg of the prescribed jig rotation. FHA intercepts were most stable when determined using trackers closest to the hinge axis. Increasing the moving window size improved the FHA direction cosines and center of rotation accuracy. Window sizes larger than 2 deg had an intercept deviation of less than 1 mm. Furthermore, compared to the 0 deg window size, the 2 deg window had a 90% improvement in FHA intercept precision while generating almost an equivalent number of FHA axes. This work identified a solution to improve FHA calculations for biomechanical researchers looking to describe changes in 3D joint motion.

Crustal structure of the Tibetan Plateau: A surface-wave study by a moving window analysis

1977 ◽  
Vol 67 (3) ◽  
pp. 735-750
Author(s):  
Kin-Yip Chun ◽  
Toshikatsu Yoshii
Keyword(s):  
Tibetan Plateau ◽  
Crustal Structure ◽  
The Tibetan Plateau ◽  
Moving Window ◽  
Low Velocity ◽  
Period Range ◽  
Window Analysis ◽  
Dispersion Data ◽  
Moving Window Analysis ◽  
Group Velocities

abstract Group velocities of fundamental-mode Rayleigh and Love waves are analyzed to construct a crustal structure of the Tibetan Plateau. A moving window analysis is employed to compute group velocities in a wide period range of 7 to 100 sec for 17 individual paths. The crustal models derived from these dispersion data indicate that under the Tibetan Plateau the total crustal thickness is about 70 km and that the crustal velocities are generally low. The low velocities are most probably caused by high temperatures. A low-velocity zone located at an intermediate depth within the crust appears to be strongly demanded by the observed dispersion data. The main features of the proposed crustal structure will place stringent constraints on future tectonic models of the Tibetan Plateau which is generally regarded as a region of active deformation due to the continent-continent collision between India and Asia.

Moving‐window Poisson analysis of gravity and magnetic data from the Penokean orogen, east‐central Minnesota

Geophysics ◽  
1991 ◽  
Vol 56 (1) ◽  
pp. 123-132 ◽  
Author(s):  
Val W. Chandler ◽  
Kelley Carlson Malek
Keyword(s):  
Magnetic Data ◽  
Superior Province ◽  
Moving Window ◽  
Thrust Belt ◽  
Early Proterozoic ◽  
Window Analysis ◽  
Fold And Thrust Belt ◽  
Gravity And Magnetic ◽  
Moving Window Analysis ◽  
Gravity And Magnetic Data

Analytical correlation of gravity and magnetic data through moving‐window application of Poisson's theorem is useful in studying the complex Precambrian geology of central Minnesota. Linear regression between the two data sets at each window position yields correlation, intercept, and slope parameters that quantitatively describe the relationship between the gravity and magnetic data and, in the case of the slope parameter, are often accurate estimates of magnetizatons‐to‐density ratios (MDR) of anomalous sources. In this study, gridded gravity and magnetic data from a 217.6 × 217.6 km area in central Minnesota were analyzed using a 8.5 × 8.5 km window. The study area includes part of the Early Proterozoic Penokean orogen and an Archean greenstone‐granite terrane of the Superior Province. The parameters derived by the moving‐window analysis show striking relationships to many geologic features, and many of the MDR estimates agree with rock property data. Inversely related gravity and magnetic anomalies are a characteristic trait of the Superior Province, but moving‐window analysis reveals that direct relationships occur locally. In the Penokean fold‐and‐thrust belt, gravity and magnetic highs over the Cuyuna range produce a prominent belt of large MDR estimates, which reflect highly deformed troughs of iron‐formation and other supracrustal rocks. This belt can be traced northeastward to sources that are buried by 3–5 km of Early Proterozoic strata in the Animikie basin. This configuration, in conjunction with recent geologic studies, indicates that the Animikie strata, which may represent foreland basin deposits associated with the Penokean orogen, unconformably overlie parts of the fold‐and‐thrust belt, and that earlier stratigraphic correlations between Cuyuna and Animikie strata are wrong. The results of this study indicate that moving‐window Poisson analysis is useful in the study of Precambrian terranes.

Human Activity Recognition using Fourier Transform Inspired Deep Learning Combination Model

2019 ◽  
Vol 9 (1) ◽  
pp. 16-31
Author(s):  
Kyungkoo Jun
Keyword(s):  
Fourier Transform ◽  
Deep Learning ◽  
Short Term Memory ◽  
Window Size ◽  
Sensor Data ◽  
Data Sets ◽  
Data Set ◽  
Proposed Model ◽  
Testing Data ◽  
Labeling Scheme

Background & Objective: This paper proposes a Fourier transform inspired method to classify human activities from time series sensor data. Methods: Our method begins by decomposing 1D input signal into 2D patterns, which is motivated by the Fourier conversion. The decomposition is helped by Long Short-Term Memory (LSTM) which captures the temporal dependency from the signal and then produces encoded sequences. The sequences, once arranged into the 2D array, can represent the fingerprints of the signals. The benefit of such transformation is that we can exploit the recent advances of the deep learning models for the image classification such as Convolutional Neural Network (CNN). Results: The proposed model, as a result, is the combination of LSTM and CNN. We evaluate the model over two data sets. For the first data set, which is more standardized than the other, our model outperforms previous works or at least equal. In the case of the second data set, we devise the schemes to generate training and testing data by changing the parameters of the window size, the sliding size, and the labeling scheme. Conclusion: The evaluation results show that the accuracy is over 95% for some cases. We also analyze the effect of the parameters on the performance.

Smoothing imaging condition for shot-profile migration

Geophysics ◽  
2007 ◽  
Vol 72 (3) ◽  
pp. S149-S154 ◽  
Author(s):  
Antoine Guitton ◽  
Alejandro Valenciano ◽  
Dimitri Bevc ◽  
Jon Claerbout
Keyword(s):  
Time Domain ◽  
Signal To Noise Ratio ◽  
Wiener Filter ◽  
Window Size ◽  
Trial And Error ◽  
Data Set ◽  
Smoothing Operator ◽  
Imaging Condition ◽  
Damping Parameter ◽  
The Time Domain

Amplitudes in shot-profile migration can be improved if the imaging condition incorporates a division (deconvolution in the time domain) of the upgoing wavefield by the downgoing wavefield. This division can be enhanced by introducing an optimal Wiener filter which assumes that the noise present in the data has a white spectrum. This assumption requires a damping parameter, related to the signal-to-noise ratio, often chosen by trial and error. In practice, the damping parameter replaces the small values of the spectrum of the downgoing wavefield and avoids division by zero. The migration results can be quite sensitive to the damping parameter, and in most applications, the upgoing and downgoing wavefields are simply multiplied. Alternatively, the division can be made stable by filling the small values of thespectrum with an average of the neighboring points. This averaging is obtained by running a smoothing operator on the spectrum of the downgoing wavefield. This operation called the smoothing imaging condition. Our results show that where the spectrum of the downgoing wavefield is high, the imaging condition with damping and smoothing yields similar results, thus correcting for illumination effects. Where the spectrum is low, the smoothing imaging condition tends to be more robust to the noise level present in the data, thus giving better images than the imaging condition with damping. In addition, our experiments indicate that the parameterization of the smoothing imaging condition, i.e., choice of window size for the smoothing operator, is easy and repeatable from one data set to another, making it a valuable addition to our imaging toolbox.

Sign in / Sign up

Export Citation Format

Share Document