Trajectory Smoothness

Signal Complexity as a Measure of Trajectory Smoothness

The measure of movement, or trajectory, smoothness is a feature that has been conventionally quantified by using higher-order time derivatives of position (i.e. jerk). Such higher-order terms lead to low signal-to-noise ratios and are difficult to validate due to noise saturation. We are investigating a new application of sample entropy (SampEn) as a measure of trajectory smoothness by applying the algorithm to the fast Fourier transform of movement speed profiles. Recent comparisons of this metric to conventional jerk-based measures has found that there is a strong correlation between the two methods. Work is ongoing to further understand if SampEn(FFT(s(t))) can be used as a reliable metric for trajectory smoothness.