End-cloud collaborative indoor electromagnetic fingerprint updating method and system

The invention provides an end-cloud collaborative indoor electromagnetic fingerprint updating method and system, and belongs to the field of indoor positioning. The method is divided into an offline stage and an online stage. In the offline process, electromagnetic fingerprints are marked according to Pearson's correlation coefficients to construct a credible fingerprint evaluation data table, and a noise reduction automatic encoder-convolutional neural network credible evaluation model is trained according to the credible fingerprint evaluation data table; in the online stage, collected fingerprint data is calculated by means of terminal indoor positioning, a credible fingerprint evaluation data table is continuously improved through normalization processing and credible electromagnetic fingerprint evaluation, affine propagation clustering analysis is carried out on a fingerprint positioning data table, and a correlation coefficient considering positions and signal attributes is calculated by using a Pearson