To address the inaccuracy of multi-storey indoor positioning, the paper proposes a method involving WiFi fingerprint positioning based on the prior information of map environment and position correction based on the image recognition technique. First, we establish a model of map environment information at the off-line stage, then identify the storey using the storey discrimination algorithm of the RSSI threshold value, and finally compute the approximate position of the terminal in combination

more »

... ith the map information model. Meanwhile, the image recognition technique is employed to correct the terminal position. Compared with the traditional fingerprint positioning method, the indoor positioning technique can not only improve the positioning accuracy but also reduce the complexity of the fingerprint matching computation at the on-line stage.