An eye movement interaction method and device based on head timing signal correction

Abstract

The invention discloses an eye movement interaction method and device based on head timing signal correction, belonging to the technical field of computer vision. The method includes: collecting a large number of continuous multi-frame binocular images and corresponding head movement timing information, actual screen gaze point coordinates, and preprocessing the collected data; using the collected large amount of data to train the gaze point of a deep convolutional network The prediction model obtains the mapping relationship between continuous multi-frame binocular images and head movement timing information to the estimated fixation point coordinates; the existing or real-time collected data is input into the trained deep convolutional network fixation point prediction model to obtain the estimated fixation point point. The invention fuses head timing signals and eye image timing signals, and uses the powerful feature processing capability of deep learning to accurately and quickly estimate the gaze point of the human eye; through pre-training the gaze point prediction model of the deep convolution network, new users do not need to correct it when using it , you can directly wear the head-mounted eye tracker for real-time eye movement interaction.

Description

technical field [0001] The present invention relates to the technical field of computer vision, in particular to an eye movement interaction method and device based on head timing signal correction. Background technique [0002] Eye movement interaction technology is to track the state of human eye gaze direction through various methods such as eye electrical signals, optical signals, and picture signals, and then use this method to conduct human-computer interaction. It has a huge role in future human-computer interaction. application potential. External devices capable of input and output and corresponding software, namely keyboards, mice, and various pattern recognition devices can complete traditional human-computer interaction, but traditional human-computer interaction is mostly manual operation, sometimes with voice operation, which has certain For example, for disabled persons with inconvenient upper limbs or pilots in military battles, traditional human-computer in...

AI Technical Summary

Problems solved by technology

[0005] In view of this, the present invention provides an eye movement interaction method and device based on head timing signal correction, the main purpose of which is to solve the problems of traditional eye movement interaction methods by integrating human eye images, eye movement timing information and head movement timing information. It is sensitive to the brightness of the environment and the degree of opening and closing of human eyes, and has poor robustness and low accuracy in complex environments.

Images