本文選題：PCA算法 + LDA算法��； 參考：《湖南大學(xué)》2016年碩士論文

【摘要】：伴隨著科技水平的進(jìn)步,管理方式的科技含量越來越高,但能否快捷、準(zhǔn)確的獲取到員工的考勤信息仍然是很多企業(yè)人力資源管理部門所急需解決的問題。而且目前,仍有少數(shù)企事業(yè)單位依然使用傳統(tǒng)的方式對員工的出勤信息進(jìn)行管理,容易出現(xiàn)"臟數(shù)據(jù)"的現(xiàn)象,因此建立一套方便、快捷且滿足企業(yè)實際的考勤管理信息系統(tǒng)逐漸成為企業(yè)事業(yè)單位考勤管理的一個勢在必行的發(fā)展方向。本文首先分析了人臉識別技術(shù)的研究背景和現(xiàn)狀,從系統(tǒng)的整體著眼,結(jié)合了企業(yè)對考勤管理的業(yè)務(wù)流程,分析了人臉識別考勤系統(tǒng)的可行性,給出了功能需求和非功能需求,同時還介紹了實現(xiàn)系統(tǒng)的關(guān)鍵技術(shù)如Adaboost算法、PCA算法和LDA算法的研究。在系統(tǒng)的設(shè)計中,采用了統(tǒng)一建模語言對系統(tǒng)進(jìn)行了總體的設(shè)計,包括系統(tǒng)的邏輯架構(gòu)、功能架構(gòu)、網(wǎng)絡(luò)架構(gòu)以及數(shù)據(jù)庫的設(shè)計,還利用流程圖和時序圖對系統(tǒng)各個功能模塊進(jìn)行了詳細(xì)設(shè)計。針對PCA算法提取特征時容易發(fā)生小樣本問題,提出了融合PCA和LDA算法來提取人臉特征,并將該算法運用到人臉識別考勤管理系統(tǒng)中。最后介紹了系統(tǒng)的實現(xiàn),系統(tǒng)采用客戶端/服務(wù)器體系結(jié)構(gòu),實現(xiàn)了對員工考勤的管理。系統(tǒng)分為三個功能模塊:考勤管理、基本信息管理和考勤統(tǒng)計管理,三個模塊之間相輔相成,實現(xiàn)了員工考勤的信息化管理。系統(tǒng)測試和運行結(jié)果顯示,系統(tǒng)界面友好、操作簡單、數(shù)據(jù)可靠,不僅能夠滿足企業(yè)考勤管理工作的需求,還優(yōu)化了企業(yè)考勤管理的流程,降低了管理成本,有效地提高了企業(yè)的管理效率,同時,嚴(yán)格的數(shù)據(jù)獲取方式與數(shù)據(jù)檢測流程保障了考勤管理系統(tǒng)中數(shù)據(jù)的完整性與安全性。
[Abstract]:With the progress of science and technology, the scientific and technological content of management mode is higher and higher, but it is still an urgent problem for many enterprise human resources management departments to get the attendance information of employees quickly and accurately. And at present, there are still a few enterprises and institutions that still use the traditional way to manage the attendance information of their employees, which is prone to the phenomenon of "dirty data", so it is convenient to establish a set. Fast and satisfying the actual attendance management information system has gradually become an imperative development direction of attendance management in enterprises and institutions. This paper first analyzes the research background and present situation of face recognition technology, from the whole system, combined with the business process of enterprise attendance management, analyzes the feasibility of face recognition attendance system, and gives the functional and non-functional requirements. At the same time, the key technologies of the system, such as Adaboost algorithm and LDA algorithm, are introduced. In the design of the system, the general design of the system is carried out by using the unified modeling language, including the logical structure, the functional architecture, the network architecture and the database design of the system. In addition, each function module of the system is designed in detail by flow chart and timing diagram. In order to solve the problem of small sample when PCA algorithm is easy to extract features, a fusion of PCA and LDA algorithms is proposed to extract face features, and the algorithm is applied to face recognition attendance management system. Finally, the implementation of the system is introduced. The system adopts the client / server architecture, and realizes the management of employee attendance. The system is divided into three functional modules: attendance management, basic information management and attendance statistics management. The three modules complement each other and realize the information management of employee attendance. The system test and running results show that the system has friendly interface, simple operation and reliable data. It can not only meet the needs of enterprise attendance management, but also optimize the process of enterprise attendance management and reduce the management cost. At the same time, strict data acquisition and data detection flow ensure the integrity and security of data in attendance management system.
【學(xué)位授予單位】：湖南大學(xué)
【學(xué)位級別】：碩士
【學(xué)位授予年份】：2016
【分類號】：TP311.52;TP391.41

【參考文獻(xiàn)】

相關(guān)期刊論文 前3條

1 秦小文;溫志芳;喬維維;;基于OpenCV的圖像處理[J];電子測試;2011年07期

2 楊利平;龔衛(wèi)國;辜小花;李偉紅;杜興;;完備鑒別保局投影人臉識別算法[J];軟件學(xué)報;2010年06期

3 田啟川;張潤生;;生物特征識別綜述[J];計算機應(yīng)用研究;2009年12期

，

本文編號：1927275