Cross-modal-mapping-based heart rate estimation method for ballistocardiogram signals

The invention provides a cross-modal-mapping-based heart rate estimation method for ballistocardiogram signals. The method is used for solving the technical problem in the prior art that the accuracyof heart rate estimation is relatively low. The method comprises the implementation steps: acquiring the ballistocardiogram signals and heart pulse signals; filtering the ballistocardiogram signals; acquiring a training sample set and a testing sample set by using periodic prior knowledge of the ballistocardiogram signals; constructing a cross-modal mapping model based on periodicity and amplitudecharacteristics of the ballistocardiogram signals; training the cross-modal mapping model; and carrying out calculating by a peak searching algorithm and an average heart rate method, thereby obtaining a heart rate estimated value of the ballistocardiogram signals. According to the method, an idea of cross-modal mapping is introduced into heart rate estimation of the ballistocardiogram signals, and the ballistocardiogram