Book Description
The purpose of this application is to develop an embedded system application which is able to collect face image and recognize the face by comparing with the database inside the system. Face recognition as a type of biometric methods has the features of non-contact, safety and convenience. It is widely used in human-computer interaction, transaction authentication, security and other fields. Recent years, with the development of mobile internet and embedded computer, it becomes possible to run face recognition on embedded system. This type of application has huge potential in remote payment and personal information security. This application is running on Android operating system which is an operating system based on the Linux kernel, and designed primarily for touchscreen mobile devices such as smartphones and tablet computers. The procedure of face recognition includes face detection, face normalization and recognition. This paper studies these key issues and successfully developed an application with nice recognition rate. The main contents and results are as follows:1) Discusses the face detection method. It used Adaboost algorithm and Haar features to detect human faces.2) Studies image pre-processing methods. Standardize the images so as to minimize the storage space and speed up the computation speed. 3) Summarize a variety of face recognition algorithms especially principle component analysis which is used in this application. Discuss the theoretical foundation of PCA algorithm. 4) Fulfilled all the features from face detection to recognition in Android platform. Using ORL face image database for testing and got a correct identification rate of over 85%. Fully verify the effectiveness of the program. Discuss the results and identification strategies.