Aerial handwriting input method based on WiFi channel state information

Abstract

The embodiment of the invention discloses an aerial handwriting input method based on WiFi channel state information, and the method comprises the following steps: obtaining a channel state information (CSI) sequence sampled along with time, and carrying out the preprocessing; obtaining an interest component amplitude and an interest component phase from the CSI sequence; comparing the interestedcomponent amplitude with an amplitude threshold determined when the finger is static, and identifying an effective input action; calculating a reflection path length change corresponding to the effective input action by utilizing the interest component phase, and calculating and obtaining hand displacement-time information of the user according to the reflection path length change and the spatialpositions of the at least two pairs of transceiving pair equipment; determining aerial handwriting input content according to the effective input action and the hand movement-time information. According to the technical scheme, more detail information is reserved by adopting a geometrical characteristic analysis mode of the complex plane track, and the precision of the displayed candidate items isimproved when the handwriting content in the air of the user is determined, so that the virtual input efficiency of the user is improved.

Description

technical field [0001] The present disclosure relates to wireless communication and radio frequency perception technology, and in particular to an air handwriting input method based on WiFi channel state information. Background technique [0002] Hands entering virtual input in the air is an important information input method in human-computer interaction, which can provide support for many applications. In the prior art, there are mainly ways to realize the air handwriting technology based on identification based on special sensor equipment and identification based on radio signals generated during normal communication. [0003] Recognition based on a dedicated sensor device allows some kind of sensor to be carried on the finger, and depending on the sensor, the way of sensing the finger's movement trajectory is also different. However, this method is expensive, and the worn device will affect the hand movement, making the hand input unnatural. [0004] The identification...

AI Technical Summary

Problems solved by technology

Direct filtering or differential interest component acquisition methods are prone to loss of detail information, mixed with the influence of non-interest components, resulting in reduced accuracy and poor robustness

Images