研究了基于智能小区安防报警系统的人脸识别方法的应用。通过仿真试验,将子图像个数、奇异值向量保留个数、聚类因数α、γ因数的选取进行了对比分析。基于聚类算法人脸识别的实际应用要求,试验数据为进一步研究子图像的赋值提供了有效的帮助和参数支持。
Application of face recognition based on security system in intelligent community was researched. The selection of sub-image's number, remaining vector' s number of singular value decomposition, clustering factor a and factor T through simulation experiments were compared and analyzed. In order to achieve requirement of practical appli- cation based on clustering algorithm for face recognition, research on weight factor of sub-image was offered effective assist and parameter support by experimental data.