人脸识别系统在嵌入式门禁系统中的研究与设计

发布时间：2018-01-30 06:22

本文关键词： 人脸识别嵌入式系统 Hi3515 预处理人脸检测特征提取　出处：《太原理工大学》2013年硕士论文　论文类型：学位论文

【摘要】：在计算机技术的快速发展过程中,生物工程技术也得到了迅猛发展。人们对身份验证实现自动化的要求逐渐提高。通过人体生物特征来进行身份识别已经成为安防系统的首选。相比其他生物特征,人脸特征具有非接触性、稳定性好、难仿冒等优点,易于为用户所接受。所以人脸识别技术具有广阔的市场应用前景。本课题通过把人脸识别技术与嵌入式技术相结合,针对人脸识系统上的具体实现问题,分别从硬件系统设计,嵌入式人脸识别软件设计以及人脸识别过程三方面对人脸识别技术在以Hi3515高性能音视频处理器为核心的嵌入式系统上的实现作了深入研究。硬件部分介绍了基于Hi3515处理器的人脸识别系统的硬件环境,主要分为重力传感模块、人脸采集模块、Hi3515中心处理模块、门禁控制模块等。嵌入式人脸识别软件设计实现部分主要包括三大部分：重力传感部分、人脸识别部分、门控部分。在计算机视觉领域里,一套完整的人脸识别系统全过程有人像获取、预处理、人脸检测、提取特征以及识别或确认等。针对人脸识别技术在嵌入式门禁系统中的研究,为了验证嵌入式人脸识别门禁系统的可靠和实时功能,提出了基于Hi3515音视频微处理器的嵌入式人脸识别门禁系统。该系统是基于Hi3515提供高性能的音频/视频服务,并采用ORL人脸库,基于以Gabor变换与TV+L’模型的图像预处理,以肤色空间(YCrCb色彩空间)的人脸检测,以主成分分析(PCA)法的特征提取的人脸识别算法在嵌入式门禁系统中得到了实现,从而根据人脸识别技术控制整个门禁系统。相应的实验结果表明,该算法在嵌入式平台上得到了优化,嵌入式系统的高可靠性和实时性问题得以解决。
[Abstract]:In the rapid development of computer technology. Bioengineering technology has also been developed rapidly. The requirement of automatic authentication has been raised gradually. Identification through human biometrics has become the first choice of security and defense system. Compared with other biometric features. . Face features have the advantages of non-contact, good stability, difficult to copy, and are easily accepted by users. So face recognition technology has a broad market application prospects. By combining face recognition technology with embedded technology, this paper designs the hardware system for the specific realization of face recognition system. The design of embedded face recognition software and the process of face recognition are studied in detail. The realization of face recognition technology in embedded system with Hi3515 high performance audio and video processor as the core. Hardware part. This paper introduces the hardware environment of face recognition system based on Hi3515 processor. It is mainly divided into gravity sensing module, face acquisition module Hi3515 center processing module, access control module and so on. The embedded face recognition software design and implementation part mainly includes three parts: gravity sensing part. In the field of computer vision, a complete face recognition system is used to obtain, preprocess, detect, extract features and recognize or confirm human face. Aiming at the research of face recognition technology in the embedded access control system, in order to verify the reliability and real-time function of the embedded face recognition system. This paper presents an embedded access control system for face recognition based on Hi3515 audio and video microprocessor, which provides high performance audio / video service based on Hi3515 and uses ORL face database. Face detection based on Gabor transform and TV L' model image preprocessing and color space YCrCb color space. The face recognition algorithm based on PCA-based principal component analysis (PCA) method is implemented in the embedded access control system, and the whole access control system is controlled according to the face recognition technology. The algorithm is optimized on embedded platform, and the high reliability and real-time problem of embedded system can be solved.
【学位授予单位】：太原理工大学
【学位级别】：硕士
【学位授予年份】：2013
【分类号】：TP391.41;TP368.1

【参考文献】