摘要
基于互联网云计算的人脸识别算法是人工智能的核心算法,也是计算机视觉发展的瓶颈所在。利用主成分分析算法对图像的特征点进行提取,并对传统的细菌觅食算法进行改进,提出了改进的细菌觅食算法(GBFA),利用改进细菌觅食算法对主成分分析法中的目标函数进行求解。将计算过程嵌入MATHEMATICA Cloud云中,利用部署函数对图片进行操作,开发人脸识别程序Face_Recognition,并分别针对智能抠图抽丝、人脸分类、同一人物人脸匹配、寻找特定人物4种功能形态进行测试,测试结果显示,提出的基于互联网云计算的人脸识别算法对于人脸识别、人脸分类、人脸筛选等具有极强的适应性和极高的精确度。
Face recognition algorithm based on Internet cloud computing is the core algorithm of artificial intelligence,and also the bottleneck lies in the development computer vision.In this paper,the features of the images have been extracted by the principal component analysis algorithm,and some improvements have been made for the traditional bacterial foraging algorithm.An improved bacterial foraging algorithm(GBFA)has been put forward.The target function in the principal component analysis has been solved by the improved bacterial foraging algorithm.The calculation process is embedded in Mathematica Cloud.By using the deployment function,the images have been manipulate,and the face recognition program has been developed.The four functional forms of smart matting spinning,face classification,face matching of the same people,and looking for a specific person have been tested.The results show that the face recognition algorithm based on the Internet cloud proposed in this article has a strong adaptability and high accuracy for face recognition,face classification,face screening,and so on.
出处
《长春工程学院学报(自然科学版)》
2016年第3期111-115,121,共6页
Journal of Changchun Institute of Technology:Natural Sciences Edition