CAJ | 학술논문

针对现有门禁系统中存在的技术缺陷、维护成本高等问题，本文设计了一套基于 ARM+Linux 的人脸识别门禁系统。该系统采用S3C2410芯片为处理器、IP网络摄像头作为视频采集设备；采用PCA算法进行人脸图像识别，并以识别输出来控制电控锁模块，从而实现人脸识别门禁系统的智能控制。最后，在不同的环境下进行了人脸识别测试，测试结果表明，该系统能准确快速高效地识别人脸，平均识别率高达99%。
침대현유문금계통중존재적기술결함、유호성본고등문제，본문설계료일투기우 ARM+Linux 적인검식별문금계통。해계통채용S3C2410심편위처리기、IP망락섭상두작위시빈채집설비；채용PCA산법진행인검도상식별，병이식별수출래공제전공쇄모괴，종이실현인검식별문금계통적지능공제。최후，재불동적배경하진행료인검식별측시，측시결과표명，해계통능준학쾌속고효지식별인검，평균식별솔고체99%。
Aiming at these problems existing in the current entrance guard system, such as technical shortcomings and expensive maintenance cost, this paper designs a face recognition access system based on ARM9 and Linux. The S3C2410 chip and IP network camera are chosen as processor and video acquisition devices respectively. Then, the facial image recognition is implemented by the PCA algorithm and the electric lock module is controlled by utilizing therecognition output, and thus an intelligent control of face recognition entrance guard system is realized. Finally, sufficient facial recognition tests are performed under different environments, and the test results show that the system can recognize faces accurately, quickly, efficiently, and the average recognition rate reaches up to 99%.