Research On Techniques Of Face Detection And Recognition For Baby Surveillance

With the development, video surveillance has become popular i-n our life. Wecan see cameras everywhere, involving some fields, such as security, human-machineinteraction, smart home, intelligent city and so on. At the time of informationexplosion, video surveillance produces a lot of video data. To meet more requirementsof people, more and more technologies and functions become part of surveillancesystem. For example, face detection, face recognition and face expression recognitionmake the system more intelligent and humanized.This paper has done some researches about face treatment technology whichmainly includes face detection, face recognition and face expression recognition. Weanalyze the development situation of the three technologies at first, and then propose afeasible implementation method for per part. For face detection, we propose a methodbased on color of skin detection and Adaboost algorithm, which is accurate andreal-time. For face recognition and face expression recognition, this paper introducesthe PCA (Pricipal Component Analysis) algorithm, and propse a method based on it.Meanwhile, we have done much about infants’ face expression recognition. Thispaper has built a infants’ face database which by we recognize four differentexpressions (normal, crying, smiling and sleepy). At last, to make sure thesetechnologies can be implemented in video surveillance, we figure out a solution basedon DM365embedded platform.