Face Detection And Tracking In Computer Video Monitoring Research

Face detection and tracking is an important problem in computer vision and face information processing.As the developing of information technology and computer technology,face detection and tracking has been widely used in many fields such as video conference,video surveillance,image and video retrieval,intelligent human-computer interaction system.It has useful value and research meaning.In this file,we focus on the research of face detection and tracking in color image sequences.On face detection,we correct the color bias based on a lighting compensation technique that adjusts the color-histogram of picture linearly by the methods of “reference white”.Then,some commonly color models in computer vision field are analysed,the YCb Cr color space system is adopted,skin color model is established by using the characteristic of aggregation of the skin color,and the detect the skin region and non-skin region by the model.Based on the pre-knowledge of human face,correct face region would be found in picture.Camshift algorithm based on color histogram is adapted to face tracking.Camshift algorithm is based on Mean Shift.It uses color histogram as tracking clues thus get rapid and accurate face tracking.But the efficency of the algorithm will be lower if there are skin color background or the tracking object move fast.To improve the algorithm's efficency,Kalman filter is introduced and combined with the Camshift algorithm.In the initial frame of video sequences,we detect the human face by using face detecting algorithm and then track the face by Camshift combined with Kalman filter.This method could solve the tracking problem under the condition of background,face masking and so on.