基于SWE的嵌入式人脸识别系统的研究与设计
首发时间:2011-04-29
摘要:人脸识别一直是国内外计算机图像处理关注的热点。本文根据扩展LBP算子、多尺度分析、等价类模式等技术实现了嵌入式人脸识别系统的设计,在以ORL标准人脸图像库中,最高取得了97.50%的识别率,可进行有效的身份鉴定。该系统将嵌入式设备与人脸识别技术相结合,充分发挥了嵌入式设备处理速度快、功耗低、易操作和网上数据交换等优势。同时采用SWE作为全新的设计模式构建应用程序,将设计与开发有效的分开,大大加快了开发速度也减少了开发错误。
关键词: 人脸识别 嵌入式 SWE 局部二值模式(LBP) WinCE
For information in English, please click here
Research and Design an Embedded System of Face Recognition Based on SWE
Abstract:Face recognition is always the focus of attention in computer image processing at home and abroad. According to expand LBP operator, multiscale analysis and equivalence class pattern technologys, the paper designs an embedded face recognition system. It achieves the highest 97.50% recognition rate in the ORL face image database,and can be an effective identity identification. The system combines embedded devices with face recognition technology, and gives full play to the embedded device processing speed, low power consumption, easy operation, online data exchange and other advantages. At the same time, it uses the SWE as the new design model to build an application which puts the design and development of effective separate, and greatly speeds up the development speed and reduces the development errors.
Keywords: Face Recognition Embedded SWE Local Binary Pattern (LBP) WinCE
基金:
论文图表:
引用
No.4423782535246130****
同行评议
共计0人参与
勘误表
基于SWE的嵌入式人脸识别系统的研究与设计
评论
全部评论0/1000