Electric Wheelchair Pedestrian Recognition and Following Technology Based on Binocular Vision
A method for electric wheelchair pedestrian recognition and following based on binocular vision is proposed to enhance the autonomy of electric wheelchairs in this paper.Firstly,the intelligent wheelchair system structure of this study is introduced.Secondly,a pedestrian skeleton key point detection method based on deep neural network is adopted,and the tracking target is re-recognized and the position is obtained by deploying a pedestrian attribute recognition model.Finally,combined with binocular vision,the local state tracking of pedestrians in the camera coordinate system is realized,allowing the wheelchair to complete the pedestrian following task in complex and changeable scenes.
intelligent wheelchairdeep learningbinocular visionpedestrian attribute recognitionperson following
万方数据
维普
