Kostas Gianikellis

A simple SSA-based de-noising technique to remove ECG interference in EMG signals

Abstract Electromyography (EMG) signals provide significant information of muscle activity that may be used, among others, to estimate the activation stages during a certain activity or to predict fatigue. Heart activity or electrocardiogram (ECG) is one of the main contamination sources, especially in trunk muscles. This paper proposes a novel method based on Singular Spectrum Analysis (SSA) and frequency analysis to separate both signals present in the raw data. The performance of the method has been compared in time and frequency domains with traditional high-pass filtering or novel techniques such as Complete Ensemble Empirical Mode Decomposition or Wavelets analysis. The results show that for both time and frequency domains the proposed approach outperforms the other methods. Thus, the proposed SSA approach is a valid method to remove the ECG artifact from the contaminated EMG signals without using an ECG reference signal.

Path Relinking Particle Filter for Human Body Pose Estimation

This paper introduces the Path Relinking Particle Filter (PRPF) algorithm for improving estimation problems in human motion capture. PRPF hybridizes both Particle Filter and Path Relinking frameworks. The proposed algorithm increases the performance of general Particle Filter by improving the quality of the estimate, by adapting computational load to problem constraints and by reducing the number of required evaluations of the weighting function. We have applied the PRPF algorithm to 2D human pose estimation. Experimental results show that PRPF drastically reduces the MSE value to obtain the set of markers with respect to Condensation and Sampling Importance Resampling (SIR) algorithms.

2D Human Tracking by Efficient Model Fitting Using a Path Relinking Particle Filter

This paper presents a 2D model-based Path Relinking Particle Filter (PRPF) algorithm for human motion tracking and analysis applications. PRPF algorithm hybridizes both Particle Filter and Path Relinking frameworks. The proposed algorithm increases the performance of general Particle Filter by improving the quality of the estimate, by adapting computational load to problem constraints and by reducing the number of required evaluations of the weighting function. A 2D human body model was used to develop tracking with PRPF. This model consists of a simple hierarchical set of articulated limbs, which is described by geometrical parameters. It is easily adaptable to the tracking application requirements. We have applied the PRPF algorithm to 2D human pose estimation in different movement tracking activities such as walking and jumping. Qualitative experimental results show that the model-based PRPF is appropriate in 2D human pose estimation studies.

Random error propagation analysis in center of pressure signal

The force measurement system unavoidably introduces noise in the output signals. The noise in the center of pressure (COP) signal is the propagation noise arise from the combination of the components used to compute it. A framework to analyze the random error in COP signals was introduced based on the “Guide to the Expression of Uncertainty of Measurement” (GUM) approach. Furthermore, the spatiotemporal resolution criterion was used as a parameter to evaluate force measurements systems under specific experimental conditions.