To improve the traditional FastICA algorithm which is sensitive to the initial weights, this paper proposes an improved FastICA algorithm based on a super relaxation factor to extract the fetal heart rate. Methods Firstly, the fetal heart rate signals in the Czech database were centralized and whitened to remove the correlation between the signals; then, a super relaxation factor was introduced into the Newton iterative algorithm to process the randomly generated initial weights, and the fetal heart rate was extracted by the improved FastICA algorithm; finally, the extraction results of the fetal heart rate signals were evaluated by visualizing the waveforms and quantitative metrics. The experimental results show that the signal-to-noise ratio of the algorithm is improved and the fetal heart rate extracted by the improved algorithm is almost free of artifacts and ECG artifacts. The improved FastICA algorithm based on the super relaxation factor relaxes the requirement of initial weights while maintaining the convergence speed, avoids the convergence imbalance, reduces the number of iterations, and can extract a clearer fetal heart rate.