Indoor heading estimation method, device, equipment and medium

Abstract

The invention provides an indoor heading estimation method and device and a medium. The method comprises the steps that IMU data are grouped to obtain local features, the local features are integrated with preset additional features, then the local features are input into a deep learning model to obtain correlation to update the additional features, and finally dimension reduction is conducted on the integrated features to obtain a heading estimation result. According to the method, data local features can be mined carefully, and specific information of different regions is captured; additional features are updated through correlation, global feature extraction is achieved, and feature representation is enriched; the dimension reduction processing can reduce the calculation complexity, reduce the training reasoning time, remove redundant information and noise, and improve the generalization and robustness of the model. Compared with the prior art, the method can better adapt to the complexity of an indoor environment, and improves the accuracy and reliability of course estimation.

Images