Multi-Target Tracking Based on Ferguson-Hough Algorithm Against Frequency Ambiguity Issue

The short-time Fourier transform (STFT) is widely used in the analysis of non-stationary signals because of its mature principle and simple implementation, as well as in the multi-target tracking based on Doppler through-wall radar (TWR). However, due to the limitation of the uncertainty principle, when the frequencies of several echo signals are close to each other, there will be aliasing in the spectrum obtained by STFT, which will seriously degrade the accuracy of multi-target tracking. To overcome this limitation, we proposed an improved Ferguson-Hough algorithm, which combines the asymmetric Ferguson parameter model with the Hough transform and can fit the instantaneous frequency curve of the echo signal in the frequency ambiguous area, thus compensating for the loss of tracking accuracy. Experiments show that the method can effectively fit the frequencies in the frequency-ambiguous region, thus making the localization results approximate to the real data, and achieving better localization accuracy than other similar algorithms.