Movement detection based on filter with symmetric impulse response

The continuous-time detection of movement is a time-domain operation. It usually requires lowpass filters ensuring small waveform distortion. To achieve small time-domain distortion, the filters approximating linear phase, such as Bessel and Gaussian filter, are often used. However, these filters have poor selectivity. Both, high selectivity and nearly linear phase are obtained by using the filters with symmetric impulse response. In this paper, we propose a simple method for the detection of movement, which uses this type of filter. We illustrate the features of the method in the measurement of the movements during biceps curl exercise. The method proved to be robust with high accuracy of the detection. Furthermore, it is suitable for application in physical rehabilitation where variety of changes in motoric functions can be detected.