Non-contact real-time eye movement identification method based on intelligent camera

Abstract

A non-contact real-time eye movement identification method based on an intelligent camera comprises the steps that (1) the intelligent camera is used for collecting face images of detected objects and stores the collected images; (2) each eye movement state template is manufactured from the multiple obtained face images according to the different detected objects; (3) official detection is started, and after the images are preprocessed simply, eye interested areas are rapidly positioned from the face images; (4) feature matching is carried out on the intersected areas, the eye movement states are sequentially matched through the least-square matching method based on grey level transformation and geometric transformation, and an accurate identification result is obtained; (5) the original images and the identification result are correlated and output to an interaction touch screen, and the current eye movement state is displayed. The method can be sued for identifying the eye movement direction and the face inclination angle, and has high real-time performance and reliability.

Description

technical field [0001] The invention relates to eye movement recognition and detection, in particular to a non-contact real-time eye movement recognition method based on an intelligent camera. Background technique [0002] Although vision is one of the five human senses, the amount of information obtained accounts for 80% to 90% of the total information of the five senses. This information affects human behavior and consciousness, and at the same time, human psychology and consciousness are also fed back in real-time in the state changes of the eyeballs. Such as blinking, changes in pupil diameter, fixation beating (moving), eye movement speed and frequency, etc. [0003] The structure of human eye movement is precise and flexible, more agile and agile than human hands. Therefore, it is of great significance to understand human psychological changes through the recognition and detection of human eyes, or to design and develop real-time interactive systems and product equip...

AI Technical Summary

Problems solved by technology

When the light intensity is insufficient, the signal-to-noise ratio of the image will decrease, which will also lead to insufficient contrast of the image; if the light is supplemented by increasing the aperture, the depth of field will become smaller, resulting in blurred images; Can make the image oversaturated and distorted

In addition, the detection accuracy is closely related to the quality of the acquired image. High-pixel images retain more eye movement information, but impose a burden on image processing

Moreover, the process of free rotation and movement of the eyes is often accompanied by unconscious head rotation, tilt, and near-far movement, resulting in deformation, grayscale changes, or even loss of eye ROIs; eye ROIs may also be detected. Occlusions such as mirror frames, hair, hands, etc., and blinking

Images