Design And Implementation Of Face Detecting And Modeling System On Mobile Devices Based On Video

As the fast development of computer technologies and the wide applications of digital products, the amount of individual digital multimedia data is increasing rapidly. And how to unearth the aboundant information contained in these data with existing multimedia technology to provide the users a novel and powerful application becomes an important problem. And as a very important medium for human to identify and express information, human face is playing a critical role in almost every aspect of our society. In recent years, the continuous development of computer graphics technologies and the uninterrupted improvement of computer hardware and software had made the face detecting and modeling technologies applied more and more widely in aspects like entertainment, authentication and public security.At the same time, the mobile Internet combines the mobile communication technology and the Internet perfectly. Both of them had become the fastest growing and the most potential industries in the world. Their growing speed is amazing and they are being recognized by more and more users. So it is easy to predict that leading the whole Internet industry, the mobile Internet industry will create a new wave of innovation which will certainly affect every aspect of our society deeply.Based on this background, this thesis proposes a solution of detecting and modeling faces in the mobile environment and designs and implements a kind of face detecting and modeling system based on mobile devices. Considering the limited computational ability of mobile devices and the high bandwidth of mobile network, this system uses the Client-Server architecture. The functions of facial data sampling and3D face model displaying are placed on the client, while the functions of face detecting and modeling are placed on the server. Then both the efficiency and accuracy of face modeling and the mobility and freedom of users can be ensured. Furthermore, this thesis proposes a solution to handle the useless and redundant face data during the pre-processing period of face modeling based on the frames’continuity assumption.This thesis first introduces the key technologies used in this system and then analyses the system demand and makes the overall design. After that, the thesis introduces the detailed design and implementation of the main function modules in this system. Then, this thesis proposes a solution to handle the useless and redundant face data. Finally, the thesis describes the testing of the whole system and makes a summary.