Efficient tracking and ego-motion recovery using gait analysis

We present a strategy based on human gait to achieve efficient tracking, recovery of ego-motion and 3-D reconstruction from an image sequence acquired by a single camera attached to a pedestrian. In the first phase, the parameters of the human gait are established by a classical frame-by-frame analysis, using an generalized least squares (GLS) technique. The gait model is non-linear, represented by a truncated Fourier series. In the second phase, this gait model is employed within a “predict–correct” framework using a maximum a posteriori, expectation-maximization (MAP-EM) strategy to obtain robust estimates of the ego-motion and scene structure, while continuously refining the gait model. Experiments on synthetic and real image sequences show that the use of the gait model results in more efficient tracking. This is demonstrated by improved matching and retention of features, and a reduction in execution time, when processing video sequences.     

Human gait recognition in canonical space using temporal templates

A system for automatic gait recognition without segmentation of particular body parts is described. Eigenspace transformation (EST) has already proved useful for several tasks including face recognition, gait analysis, etc; it is optimal in dimensionality reduction by maximising the total scatter of all classes but is not optimal for class separability. A statistical approach that combines EST with canonical space transformation (CST) is proposed for gait recognition using temporal templates from a gait sequence as features. This method can be used to reduce data dimensionality and to optimise the class separability of different gait sequences simultaneously. Incorporating temporal information from optical-flow changes between two consecutive spatial templates, each temporal template extracted from computation of optical flow is projected from a high-dimensional image space to a single point in a low-dimensional canonical space. Using template matching, recognition of human gait becomes much faster and simpler in this new space. As such, the combination of EST and CST is shown to be of considerable potential in an emerging new biometric     

Simulating pathological gait using the enhanced linear inverted pendulum model

In this paper, we propose a new method to simulate human gait motion when muscles are weakened. The method is based on the enhanced version of three-dimensional linear inverted pendulum model that is used for generation of gait in robotics. After the normal gait motion is generated by setting the initial posture and the parameters that decide the trajectories of the center of mass and angular momentum, the muscle to be weakened is specified. By minimizing an objective function based on the force exerted by the specified muscle during the motion, the set of parameters that represent the pathological gait was calculated. Since the number of parameters to describe the motion is small in our method, the optimization process converges much more quickly than in previous methods. The effects of weakening the gluteus medialis, the gluteus maximus, and vastus were analyzed. Important similarities were noted when comparing the predicted pendulum motion with data obtained from an actual patient.     

Abnormal gait recognition using exemplar based algorithm in healthcare applications

In healthcare applications, gait plays a major role in identification of the normal or abnormal person in different situations. Human gait refers to the walking style of the person, and it may also refer as locomotion using human limbs. The abnormal gait has irregular patterns of stance and swing phases. Without any clinical impairment, this paper proposes a novel approach to classify the person as normal or the person as suffering from neurological disorders from the videos using their gait videos. In addition, neurological gait disorders such as Parkinson gait, hemiplegic gait, and neuropathic gait has been identified using the gait features. Many systems are designed to detect and identify gait disorders using head, hip, heel, and toe behavior analysis from the bidirectional gait videos. As motivated by previous mechanisms, this paper proposes a novel vision based algorithm to recognize the gait abnormalities using model free approaches and significant feature vector generation from complete silhouette images of one gait cycle of a person. Here, a lean angle and ramp angle are considered as distinguishing and prominent features, and the results of these features are properly classified into normal or abnormal gait through the design of an unsupervised classifier.     

Elevating pain thresholds in humans using depolarizing prepulses

Electrocutaneous stimulation is a potentially useful communication tool for applications in virtual reality, sensory substitution, and sensory augmentation. Many of these applications require the use of arrays of small electrodes. Stimulation through small electrodes is often painful, however, limiting the practicality of such arrays. The purpose of this study was to test a method for elevating the pain threshold to electrocutaneous stimulation through small (1-mm diameter) electrodes on the fingertip. We hypothesized that long, subthreshold, depolarizing prepulses (PP) would elevate the pain threshold so that a subsequent stimulus pulse (SP) would be less likely to be painful. We used psychophysical methods to measure the probability that an SP would be perceived as painful both by itself and when preceded by a PP that was 2, 4, 6, 8, or 10 dB lower in amplitude than the SP. We found that the PPs significantly increased the pain threshold, reducing the likelihood that the SP was painful (p < .0001). The dose effect of PP amplitude was also highly significant (p < .0001), with larger PPs elevating pain thresholds more. To our knowledge, this is the first report of PPs being used to elevate electrical stimulation thresholds in humans. PPs may be useful for selective inactivation of neural subpopulations in many human neuroprosthetic applications.     

Reducing covariate factors of gait recognition using feature selection and dictionary-based sparse coding

Human gait recognition is a behavioral biometrics method that aims to determine the identity of individuals through the manner and style of their distinctive walk. It is still a very challenging problem because natural human gait is affected by many covariate factors such as changes in the clothing, variations in viewing angle, and changes in carrying condition. This paper evaluates the most important features of gait under the carrying and clothing conditions. We find that the intra-class variations of the features that remain static during the gait cycle affect the recognition accuracy adversely. Thus, we introduce an effective and robust feature selection method based on the gait energy image. The new gait representation is less sensitive to these covariate factors. We also propose an augmentation technique to overcome some of the problems associated with the intra-class gait fluctuations, as well as if the amount of the training data is relatively small. Finally, we use dictionary learning with sparse coding and linear discriminant analysis to seek the best discriminative data representation before feeding it to the Nearest Centroid classifier. When our method is applied on the large CASIA-B gait data set, we are able to outperform existing gait methods by achieving the highest average result.     

Transmission-line electric field induction in humans using chargesimulation method

The authors discuss determining the distribution of the fields, induced charges, and currents in a human body standing in the high electric fields produced by high voltage overhead transmission lines. The authors' method of analysis is based on the charge simulation technique. This serves to explain the biological studies of possible long-term exposure effects to electric fields     

Identification of cubically nonlinear systems using undersampleddata

A practical technique for identification of cubically nonlinear systems using higher order spectra of the discrete data samples of the system input and output is proposed. This technique differs from the conventional one in that it only requires the sampling frequency for the system output to be equal to twice the bandwidth of the system input, instead of six times the bandwidth of the system input. This means the demand for high speed processing and a large amount of data in the conventional approach can be greatly relieved, Two methods are developed: one is suitable for systems with a Gaussian random input, the other is suitable for systems with a non-Gaussian random input. The advantages of the two methods over their conventional counterparts are demonstrated via computer simulation     

Identification of linear systems using correlation data

Powerful algorithms exist for identifying linear systems given input/output data. In many cases, however, it is convenient to obtain the data in the form of correlation functions. The data in this form are often more condensed. An extension of ?str?m's identification algorithm for making it able to use correlation functions in the place of an input/output record is described.     

Individualized short-term core temperature prediction in humans using biomathematical models

This study compares and contrasts the ability of three different mathematical modeling techniques to predict individual-specific body core temperature variations during physical activity. The techniques include a first-principles, physiology-based (SCENARIO) model, a purely data-driven model, and a hybrid model that combines first-principles and data-driven components to provide an early, short-term (20-30 min ahead) warning of an impending heat injury. Their performance is investigated using two distinct datasets, a Field study and a Laboratory study. The results indicate that, for up to a 30 min prediction horizon, the purely data-driven model is the most accurate technique, followed by the hybrid. For this prediction horizon, the first-principles SCENARIO model produces root mean square prediction errors that are twice as large as those obtained with the other two techniques. Another important finding is that, if properly regularized and developed with representative data, data-driven and hybrid models can be made "portable" from individual to individual and across studies, thus significantly reducing the need for collecting developmental data and constructing and tuning individual-specific models.     

Noninvasive feature-based detection of delayed gastric emptying in humans using neural networks

Radioscintigraphy is currently the gold standard for gastric emptying test which involves radiation exposure and is considerably expensive. We present a feature-based detection approach using neural networks for the noninvasive diagnosis of delayed gastric emptying from the cutaneous electrogastrogram (EGG). Simultaneous recordings of the EGG and scintigraphic gastric emptying test were made in 152 patients with symptoms suggestive of delayed gastric emptying. Spectral analyses were performed to derive EGG parameters which were used as the input of the neural network. The result of scintigraphic gastric emptying was used as the gold standard for the training and testing of the neural network. A correct classification of 85% (a specificity of 89% and a sensitivity of 82%) was achieved using the proposed method.     

Identification of nonlinear systems using the Wiener model

An algorithm for the identification of nonlinear systems which can be described by a Wiener model consisting of a linear system followed by a single-valued nonlinearity is presented. Crossconolation techniques are employed to decouple the identification of the linear dynamics from the characterisation of the nonlinear element.     

Pneumatic active gait orthosis

The paper describes design, construction and experimental testing of an active gait orthosis intended to assist locomotion in paraplegic subjects. Design specifications were formulated on the basis of an analysis of the context in which the device will be used. Different joint activation solutions were then proposed and examined. Considerable care was devoted to defining the logic and control system and to implementing the electro-pneumatic circuit, given that the system must be worn by a disabled subject who has no sensory perception in the lower limbs. Experimental tests, which made it possible to determine system performance at each step of development, were carried out with no user, then with a healthy user and finally with a paraplegic user. Experimental results consisting of graphs and photographic images are presented and discussed.     

Identification of linear time-invariant system using feedback gain

Feedback property and exponential curve fitting are used to identify the parameters of transfer function. It is assumed that all state variables are inaccessible for measurement.     

Identification of dispersive communication channels using Barker codes

The identification of dispersive sommunication channels using short duration binary signals in considered in this paper. It is shown that Barker codes are the best known finite length binary test signals for the identification of the impulse response of base-bond channels using the cross-correlation technique. The identification of RF passband channels is performed by using RF pulse with Barker code phase modulation as test signals. The described identification method is suitable for adaptive digital communication systems operating in time varying channels. It is also suitable for channels that may show non-linear effects when long duration test signals are used for identification.     

Application of a neuro-fuzzy network for gait event detection using electromyography in the child with cerebral palsy

An adaptive neuro-fuzzy inference system (ANFIS) with a supervisory control system (SCS) was used to predict the occurrence of gait events using the electromyographic (EMG) activity of lower extremity muscles in the child with cerebral palsy (CP). This is anticipated to form the basis of a control algorithm for the application of electrical stimulation (ES) to leg or ankle muscles in an attempt to improve walking ability. Either surface or percutaneous intramuscular electrodes were used to record the muscle activity from the quadriceps muscles, with concurrent recording of the gait cycle performed using a VICON motion analysis system for validation of the ANFIS with SCS. Using one EMG signal and its derivative from each leg as its inputs, the ANFIS with SCS was able to predict all gait events in seven out of the eight children, with an average absolute time differential between the VICON recording and the ANFIS prediction of less than 30 ms. Overall accuracy in predicting gait events ranged from 98.6% to 95.3% (root mean-squared error between 0.7 and 1.5). Application of the ANFIS with the SCS to the prediction of gait events using EMG data collected two months after the initial data demonstrated comparable results, with no significant differences between gait event detection times. The accuracy rate and robustness of the ANFIS with SCS with two EMG signals suggests its applicability to ES control.     

Identification of metallic mine-like objects using low frequencymagnetic fields

This paper addresses the issue of identifying conducting objects based on their response to low frequency magnetic fields; an area of research referred to as magnetic singularity identification (MSI). Real-time identification was carried out on several simple geometries. The low frequency transfer function of these objects was measured for both cardinal and arbitrary orientations of the magnetic field with respect to the planes of symmetry of the objects (i.e., different polarizations). Distinct negative real axis poles (singularities) associated with each object form the basis for their real-time identification algorithm. Recognizing this identification problem as one of inference from incomplete information, a generalized likelihood ratio test (GLRT) is presented as a solution to the M-ary hypothesis testing problem of interest. Best performance of their GLRT classification scheme, measured through Monte Carlo simulation and presented in terms of percent correct identification versus SNR, was obtained with a single pole per object orientation     

Identification of pneumatic cylinder friction parameters using genetic algorithms

A method for identifying friction parameters of pneumatic actuator systems is developed in this paper, based on genetic algorithms (GA). The statistical expectation of mean-squared errors is traditionally used to form evaluation functions in general optimization problems using GA. However, it has been found that, sometimes, this type of evaluation function does not lead the algorithms to have a satisfactory convergence, that is, the algorithm takes a long period of time or fails to reach the values of parameters to be identified. Different evaluation functions are, therefore, studied in the paper and two types of evaluation functions are found to have the expected rate of convergence and the precision. The algorithm is initially developed and tested using the benchmark data generated by simulations before it is applied for parameter identification using the data obtained from the real system measurement. The results obtained in the paper can provide the manufacturers with the observation to the characteristics inside the pneumatic cylinders.