Font Size: a A A

Research Of Embedded Face Recognition Algorithms Based On ARM

Posted on:2012-11-13Degree:MasterType:Thesis
Country:ChinaCandidate:S M WangFull Text:PDF
GTID:2218330338968515Subject:Computer Science and Technology
Abstract/Summary:PDF Full Text Request
Since the 19th century, along with the development of science and technology, face recognition technology has been developing rapidly. Especially face as a very important part of biological features, more and more people pay their attention on it. Embedded equipment especially ARM is a new product, as one of easy to carry, quick installation, mobile AFR, it has been involved in various fields, and has been widely applied in all kinds of electronic products. The design of embedded system includes the design of embedded development platform, the development of operation system, the development of system drive, the development of application etc.Based on ARM, which is a new embedded device, the recognition algorithms have been researched in the paper. The embedded operation system is first developed and transplanted, with commonly mature face recognition algorithm as the theoretical basis, the feasibility analysis and disfigurement was got, and then optimization and solution was proposed, at last embedded face recognition framework was proposed, and real face images was used on the test. Based on the ARM board embedded face recognition algorithm was implemented and the result is good. This paper mainly has done work as follows: Based on the research of Haar-like features of face and Adaboost algorithm, the face can be rapid detect based on the ARM device and camera; Based on the research of 2DPCA image dimension reduction algorithm and nearest neighbor algorithm, the embedded face recognition system can be implemented ; By introducing the two-dimensional Gabor transform algorithm, it can solve the problem of face image mode with the change of illumination conditions in face recognition system, and reduce the image noise, which is good for many posture face recognition system; Because multi-scale Gabor filter has brought high dimensional data problems, the multi-scale Gabor weighted fusion algorithm was put forward, it can well solve high dimensional data problem; Based on ARM the design process of embedded face recognition system is introduced in detail, and the working process of the system was given in detail, at last the actual image were tested, the results indicate that the recognition performance of the system is good. Because the resources of ARM board are scarce, with the improved algorithm,which is proposed in this paper, face recognition classifier has been first trained on the PC, and then through the network the classifier was transmitted to embedded ARM development platform, at last the embedded face recognition was implemented, and the actual images were tested , the results indicate that the recognition performance of the system is good. The proposed face recognition algorithm not only has certain theoretical reference value, but also for the embedded system application development has certain significance.
Keywords/Search Tags:Embedded face recognition, ARM, 2DPCA, Two-dimensional Gabor, Haar-like face detection
PDF Full Text Request
Related items