Human Eye Recognition Tracking Based on Region Segmentation and Template Matching

Abstract

As a very challenging field, computer vision has always attracted the in-depth exploration and research of many researchers because of its huge development potential. With the continuous improvement of computer performance and the popularity of electronic products, more and more researchers are committed to the important research direction in the field of computer vision - human eye detection. In this paper, the eye in the video sequence is taken as the research object, and the tracking and recognition algorithm of the dynamic object is taken as the main research purpose. The motion tracking of the human eye is deeply studied. Aiming at the existing problems of human eye tracking and recognition algorithms, this paper proposes a human eye tracking and recognition algorithm based on combination of region segmentation and template matching. Experiments show that this method greatly improves the accuracy and stability of human eye recognition.