Objective Heart sound segmentation is the basis of heart sound analysis. Most traditional methods are based on the determination of the locations and types of fundamental heart sound (FHS). Misrecognized or unrecognized problems may happen due to the interference of murmurs. This paper proposes a segmentation method based on the cardiac cycle extraction to avoid the identification of the FHS. Methods A new method based on LabVIEW performs noise reduction for raw heart sound signals by using wavelet transform, envelope detection by using fast Hilbert transform, calculation of cardiac cycle lengths by using auto-correlation of envelope signals. The lengths of cardiac cycles are used to crop a few whole cycles of denoised heart sound signals to avoid identifying the FHS. Results The proposed method is tested on 30 heart sound signals, and the results can be displayed visually with 98% accuracy. Conclusions This is a method without the need to label individual FHS, which lays a solid foundation for the subsequent research.