Research on Indoor Positioning Algorithm Based on Information Fusion

In order to make full use of the advantages brought by information fusion to improve the accuracy of indoor positioning, this paper proposes an indoor positioning algorithm based on Wi-Fi and video. The algorithm uses the received signal strength RSSI and video data as basic information. First, the model of the convolutional neural network is used to complete the fingerprint Wi-Fi indoor positioning based on RSSI and the YOLOv3 algorithm is used to detect moving targets at the same time in the same measured space. Then, based on the indoor positioning results of Wi-Fi, the indoor positioning results are transformed into the pixel coordinate system of moving object detection of the surveillance video at the same time through coordinate conversion, and the two types of target detection information are merged to complete the double positioning of moving objects. The experimental results show that under the premise that the timeliness can meet the requirements, the combination of the two methods improves the accuracy and location distribution.