CAJ | 학술논문

针对现有视频监控系统过分依赖人眼检测和缺乏智能性的弊端，根据南京师范大学敬文图书馆阅览室管理系统的要求，提出了一种基于主成分分析与AreGIS网络分析算法的目标检测、识别和显示技术，应用背景减除法分割人头轮廓，奇异值分解训练样本空间，经过PCA变换得到目标图像的特征向量并进行目标匹配；同时建立网络模型提供座位信息查询和路径导航，并通过AO对象建模和应用软件功能控制模块实现图书馆阅览室的管理和应用功能，保证图书馆信息安全和稳定运行．实验结果表明，该系统具有较好的识别效果，对图书馆现代化、智能化、网络化和人性化建设起到积极作用．
침대현유시빈감공계통과분의뢰인안검측화결핍지능성적폐단，근거남경사범대학경문도서관열람실관리계통적요구，제출료일충기우주성분분석여AreGIS망락분석산법적목표검측、식별화현시기술，응용배경감제법분할인두륜곽，기이치분해훈련양본공간，경과PCA변환득도목표도상적특정향량병진행목표필배；동시건립망락모형제공좌위신식사순화로경도항，병통과AO대상건모화응용연건공능공제모괴실현도서관열람실적관리화응용공능，보증도서관신식안전화은정운행．실험결과표명，해계통구유교호적식별효과，대도서관현대화、지능화、망락화화인성화건설기도적겁작용．
Aiming at the disadvantages of the current video monitoring system＇ s lacking intelligence and over-reliance on eye examination and identification, a new library reading room management technique is proposed for the application of Jingwen Library in Nanjing Normal University according to the libarary＇s requirements. Based on the principal component analysis （PCA） algorithm and network analysis algorithm model in ArcGISTM, target detection and identification as well as visualization techniques are performed. Head outline segmentation is firstly carried out with the background subtrac- tion method and sample training is conducted with Singular value decomposition algorithm. After the PCA transform, tar- get match is achieved for the target image with the calculated feature vectors. The seating information query and naviga- tion system is then set up by using the network model and the application management is built by using the AO function control and application software modules to guarantee information security and stable operation for library. The experi- mental results prove that the proposed system has good seating identification accuracy and suggests an active role in modernization, intelligentization and humanization for network library reading room management.