【作者】 孙超；

【导师】 贠卫国；

【作者基本信息】 西安建筑科技大学 ， 计算机应用技术， 2010， 硕士

【摘要】 随着计算机电子技术的飞速发展,市场上出现了多种多样的汽车防盗系统。本文运用ARM(Advanced RISC Machines)处理器设计了一个针对使用汽车万能钥匙盗窃汽车的防盗系统。此系统由人脸识别系统、GSM网络、步进电机以及ARM控制系统构成。本文主要内容如下：(1)对硬件系统进行了总体规划,根据设计方案对ARM开发板、USB摄像机、GSM模块步、进电机模块进行了选型,并且使用Altium designer电路设计软件完成扩展电路设计。(2)嵌入式Linux操作系统的移植。在对嵌入式Linux操作系统进行研究的基础上建立了嵌入式交叉编译环境、在分析系统引导过程后对Bootloader进行移植、使用busybox工具构建cramfs文件系统、并实现了Linux内核的编译与移植。(3)GSM和USB摄像头程序的设计与开发。首先分析了内核驱动程序构架,然后依据该架构,参考相应的硬件数据手册,实现了USB摄像机、GSM设备驱动的加载；通过Linux提供的Video4Linux模块完成了图像采集程序的编写,针对西门子TC35 GSM模块提供的AT指令集完成短消息收发程序的编写；并且完成步进电机的控制。(4)人脸识别算法的研究。首先对人脸识别技术进行概要性的介绍,对图像进行增强和归一化等预处理,用小波变换和肤色分割的的方法实现人脸的检测和定位,然后用主成分分析PCA (Principal Component Analysis)算法进行对图像人脸进行特征值提取,最后用最邻近距离分类器衡量特征匹配实现人脸的识别。整个人脸识别程序使用Intel提供的跨平台的Opencv函数库完成,在最后通过实验的方法验证人脸识别的准确性。本系统的创新点在于利用人脸识别方法判别当前驾车人身份,再通过车主手机与ARM控制器进行交互的启动方式来进行汽车的防盗,在本文结束后对全文进行了总结,对系统中存在的不足之处提出了改进性的建议。更多还原

【Abstract】 With the rapid development of computer electronic technology, a wide variety of car alarm systems appearing on the market. This papers uses an ARM processor design a car alarm system for the people who stolen cars by copy the key of the cars. This system consists of face recognition systems, GSM network, stepping motor and ARM control system.The Main content of this papers as follows:(1).Hardware system for the master plan, according to the design on the ARM Development Board, select USB video camera、GSM module and Stepping Motor, then using Altium designer circuit design software complete interface circuit design.(2).Transplant the Linux OS on Embedded system. After study the Principle of Linux OS on the embedded system, design a cross compilation environment. Analysis of the system boot process,then complete the transplantation of bootloader, use the busybox tool to build a cramfs filesystem, and complete the Linux kernel compilation and transplantation.(3).Design software for GSM and USB camera. First analysis of the kernel driver framework, and then based on the architecture, hardware reference to the corresponding data sheet, to achieve a USB camera, GSM device driver loading; Video4Linux modules provided by Linux to complete the code of image acquisition;complete the preparation of short messaging program by the Siemens TC35 GSM module provides the AT commands; and complete the control of stepper motor.(4).Research the face recognition algorithm. Firstly, introduct the face recognition technology, the image enhancement and normalization, use the wavelet theory and skin color segmentationd for face detection and location, then principal component analysis (PCA) algorithm for human face image feature extraction, and finally use the nearest distance classifier, feature matching for face recognition. Completed the face recognition program by uses cross-platform Opencv library which provide by Intel Verify the accuracy of face recognition by the method of experimental.This system of innovation is use of face recognition identifying current driver identity, Through the mobile phone interact with arm controller starts to prevent stealing cars, then summary the paper, give the suggestion for deficiencies in papers.更多还原