Sleep monitoring system and method based on electrocardiosignal

The invention provides an electrocardiosignal-based sleep monitoring system and method, which comprises the following steps of: decomposing and reconstructing an acquired electrocardiosignal, and extracting a time domain feature, a frequency domain feature and a coupling feature from a reconstruction result; generating a plurality of Hilbert spectrums according to the features, inputting the plurality of Hilbert spectrums and the apnea type as a label into a convolutional neural network model for training, thereby classifying sleep staging stages and respiratory events corresponding to each stage; in the training process in a convolutional neural network model, action information sensed by an action sensor, the sleep staging stages and the breathing time serve as labels for training, and finally the multi-dimensional multi-scale sleep analysis report is generated. The system has the advantages that the physical state of a user can be integrally monitored, and timely screening and early warning can be carried