Research On Elderly Care Needs Recognition System Based On Eye Movement And Head Movement Data Fusion Technolog

For the bedridden/semi-disabled elderly,long-term care is needed to solve their life and survival obstacles.The prerequisite for effective care is to accurately and real-time identify the care needs.In this paper,the gaze tracking technology that integrates eye movement and head movement data is used in the recognition of the care needs of the bedridden elderly.Using the gaze of the elderly to identify care needs can make up for the limitation that the elderly cannot express their needs with voice,gestures,etc.Then this paper applied this technology to the intelligent nursing bed control system independently developed by the team.The natural and convenient integrated care is provided for the elderly.The main research work of this paper includes:(1)Eye Movement Feature Parameter Extraction.The Adaboost is used to detect the face of the video image collected by the camera,and the Meanshift is used to track the face to obtain a stable facial area.According to the stable geometric relationship,the human eye area is roughly located,and the iris operator is used to further locate the human eye.Then the OTSU threshold segmentation method is used to roughly locate the pupil area,and the pixel-level and sub-pixel-level extraction of the pupil edge points are successively used to improve the accuracy.The corneal reflection area is located according to the gray value,and then the minimum circumscribed circle center of the reflection area indicates the center of the Purchin spot,realizing accurate positioning of the Purchin spot.(2)Head Pose Estimation.The paper constructs a characteristic triangle face plane with the left and right inner corners of the eyes and the midpoint of the mouth as vertices.Firstly,the binocular vision camera is stereo-calibrated to obtain the internal and external parameters,and then Harris corner detection is used to locate the feature points.According to the principle of binocular stereo ranging,the three-dimensional coordinates of the vertices of the face plane feature triangles are calculated.Calculate the angle between the face plane and the X and Y axes of the world coordinate system,which are the pitch angle and the yaw angle of the head.Experiments show that the algorithm can effectively obtain the head pose.(3)Sight Point Estimation.Combining eye movement parameters and head movement posture,a method of gaze point estimation based on the 3D model of the human eye is proposed.The optical axis is calculated using the pupil center,the light source,and the virtual image of the light source.The head movement parameters are integrated into the calibration process of the rotation matrix,and then the visual axis is calculated.At last this paper calculated the position of the gaze point according to the gaze direction of the fusion of eye movement and head movement data to verify the feasibility.(4)Application of Care Need Recognition System.The paper applied the gaze tracking algorithm to the nursing bed control system,and transmit the sight point recognition result to the main control chip of the nursing bed through the Wi Fi communication module.Realize the control of the nursing bed,and respond to the care needs of the elderly in bed.The paper provides a convenient and natural way of interaction for the disabled elderly.