Somatic action identifying method and man-machine interaction device

Abstract

The invention provides a somatic action identifying method. The somatic action identifying method includes steps of 1) acquiring effective data segments of the somatic action by a speed sensor in the somatic device; 2) calculating DTW distance between the acquired effective data segments and the standard data segments and matching a sample library according to the DTW distance between the acquired effective data segments and the standard data segments; 3) calculating the DTW distance between the acquired effective data segments and each sample data segment in the matched sample library of the step 2), and using identifiers of the somatic action indicated by the minimum sample data segment of the DTW distance as the identifying results of the acquired effective data segments. The invention further provides a corresponding man-machine interaction device. On the premise of guaranteeing identifying accuracy, identifying efficiency is improved; high identifying accuracy is guaranteed on the premise that holding ways of the device during action identification are not limited.

Description

technical field [0001] The present invention relates to the technical field of somatosensory pattern recognition, and in particular, the present invention relates to a somatosensory action recognition method and a human-computer interaction device. Background technique [0002] Speaking of the way of human-computer interaction, we first think of the way of interaction based on keyboard and mouse, but with the emergence of smart TVs and large-size display screens, the inconvenience of keyboard and mouse is becoming more and more obvious, while touch screen interaction and voice interaction become more and more inconvenient. , action-based (such as gesture-based) interaction and other new generation of interaction methods came into being. [0003] At present, there are two mainstreams of action-based interaction. One is the computer vision-based interaction mode, which is represented by Microsoft's Kinect for Xbox. Kinect has two problems. First, there is the problem of occlu...

AI Technical Summary

Problems solved by technology

However, Wii's motion recognition is limited to games, and its algorithm has not been made public, so it is difficult to transfer to other devices

In other publicly available action recognition algorithms based on wearable sensors, there are generally the following problems: one is that a large amount of data is required to train an action model, and it is not easy to increase the types of actions; The way the device takes is limited when recognizing actions. Fourth, when there are too many samples in the sample library, it is necessary to traverse all samples, resulting in low efficiency.

Images