Wearable eye tracker first perspective video image analysis method

Abstract

The invention relates to a wearable eye tracker first perspective video image analysis method. The method comprises steps of data preprocessing in which a relation graph between eye focus point coordinates and first perspective video image frames is drawn, video image frames corresponding to hopping points are found out, and video image frames corresponding to hopping points are screened out in original data of the first perspective video image frames and data analysis in which analysis is carried out according to the preprocessed first perspective video image frames. The data analysis step comprises specific steps: S1, camera calibration is carried out, and the first perspective video image is corrected; S2, image segmentation is carried out, and a target area on which the tested in the image focuses is detected and extracted; and S3, object recognition is carried out, and the target area is effectively recognized. The target area on which the tested focuses can be effectively detected and recognized, and the focus point position is automatically marked.

Description

technical field [0001] The content of the present invention belongs to the field of eye tracker data analysis, and more specifically, it is a first-view video image analysis method of a wearable eye tracker. Background technique [0002] With the development of eye-tracking technology, the development of eye-tracking instrument is becoming more and more mature, and it has good application prospects in various aspects such as psychology, medicine, usability testing, and market research. However, the analysis of eye movement data is mostly focused on the extraction and processing of fixation duration, number of fixation points, areas of interest, and gaze trajectory maps. Currently, the analysis of eye movement data is based on the existing The data analysis module usually mainly obtains visualization results such as heat maps and scanning paths. And its eye movement analysis is mainly aimed at the desktop eye tracker. The desktop eye tracker collects the eye movement data wh...

AI Technical Summary

Problems solved by technology

However, in a real scene, when wearing a wearable eye tracker, the target area that the subject pays attention to changes in real time and cannot be predicted, so it is impossible to pre-define the area of ​​interest. To further evaluate the eye movement, study a certain human For some visual behaviors and psychological cognition, it is generally necessary to manually mark the position of the subject's fixation point, which is costly and time-consuming

Images