计算机与现代化
計算機與現代化
계산궤여현대화
COMPUTER AND MODERNIZATION
2015年
6期
122-126
,共5页
人脸识别%人脸检测%CUDA%实时
人臉識彆%人臉檢測%CUDA%實時
인검식별%인검검측%CUDA%실시
face recognition system%face detection%Compute Unified Device Architecture ( CUDA)%real-time
通过对人脸识别系统的2个关键部分的优化,实现一种快速高效的人脸识别系统。在面部检测阶段改进图像积分的并行算法；在面部识别阶段尝试算法的并行化,并且把测试阶段的一部分进行了并行化。与传统的CPU识别程序相比,CUDA平台改进程序可在面部检测阶段实现22.42倍的加速比,在面部识别阶段实现1668.56倍的加速比。实验数据表明,本文提出的人脸识别系统具有很高的实时性能。
通過對人臉識彆繫統的2箇關鍵部分的優化,實現一種快速高效的人臉識彆繫統。在麵部檢測階段改進圖像積分的併行算法；在麵部識彆階段嘗試算法的併行化,併且把測試階段的一部分進行瞭併行化。與傳統的CPU識彆程序相比,CUDA平檯改進程序可在麵部檢測階段實現22.42倍的加速比,在麵部識彆階段實現1668.56倍的加速比。實驗數據錶明,本文提齣的人臉識彆繫統具有很高的實時性能。
통과대인검식별계통적2개관건부분적우화,실현일충쾌속고효적인검식별계통。재면부검측계단개진도상적분적병행산법；재면부식별계단상시산법적병행화,병차파측시계단적일부분진행료병행화。여전통적CPU식별정서상비,CUDA평태개진정서가재면부검측계단실현22.42배적가속비,재면부식별계단실현1668.56배적가속비。실험수거표명,본문제출적인검식별계통구유흔고적실시성능。
Through the optimization of the two important part of face recognition system, we proposed a CUDA-based real-time face recognition system. In the face detection phase with Viola-Jones cascade classifier, we implemented and improved novel par-allel methodologies of image integral, calculation scan window processing and the amplification and correction of classifiers. In the face recognition phase, we explored the parallelizing of the algorithm and parallelized some part of the testing phase. In compari-son with traditional CPU program, the proposed approach running on an NVIDIA GTX 570 graphics card could respectively a-chieve 22. 42 times speedup in detection phase and 1668. 56 times speedup in recognition phase. The recognition speed will in-crease until it reaches the hardware resource limit. The experimental results demonstrate that the system achieves a good real-time performance.
