透過您的圖書館登入
IP:3.133.113.101
  • 學位論文

利用人物辨別完成數位看板與推薦系統之實現

Implementation of Digital Signage and Recommended System by Using Characters Identification

指導教授 : 陳自強
若您是本文的作者,可授權文章由華藝線上圖書館中協助推廣。

摘要


隨著科技不斷的演進,人們的生活品質也隨而提升,生活中許多小細節漸漸的被新科技取代,為了貼近人心也使產品更加實用,個人化的商品也隨之誕生,本研究希望透過辨識每人獨一無二的人臉,進而獲取個人過去的資料以及其喜好的分析,達成個人化的推薦系統,也將以結合智慧收銀機與數位電子看板的概念去發展,希望可以達到研發出一種全新的客製化推播系統,讓操作者與顧客都可以獲得最好的消費體驗。 本研究由Toshiba公司的TZ3000嵌入式系統為接收資訊端,TZ3000內含Cortex - A9 CPU,並搭載環景攝影、人臉偵測、情緒分析等功能,得到這些資訊資後,再由本研究的智慧收銀機擔任接收端,來執行人臉辨識,以及資料的控管,接著將個人資料進行分析,輸出至數位電子看板來作客製化的分析與廣告推薦,而其中也對於OpenCV中人臉辨識方法只能進行預測一個最佳解的函式去進行修改,使其可傳送多個近似解,以輔助選取的方式來提升準確率,最後以情緒、年齡、性別來推播一個屬於顧客關注度最高的廣告,並記錄當下滿意度等資訊作為參考依據,實現個人化、消費、廣告結合的功能。

並列摘要


With the advance of the technology, people’s quality in lives are also progressed. Many parts in our lives are replaced with new technology gradually. It is not only close to human’s heart but also convenient for us. As a result, personal products are produced during recent year. in this thesis, we hope to get people’s information in the past and their favorites by recognizing human face in order to perform personal recommendation system. In addition, we combine the concept of smart cash register with digital electronic board to develop a whole new customized recommendation system. It can make users and customers to acquire the best consuming experience. We use the embedded system, TZ3000 which is designed by Toshiba as a receiver. The TZ3000 included core Contex-A9, equipped with panorama, face detection and emotion analysis etc. The smart cash register plays a receiver role to execute face recognition and data control and management, and then we analyze that personal data. Afterwards, it will output to digital electronic board to make customized analysis and advertisement recommendation. According to OpenCV, the method of face recognition can only predict a better solution of function. However, we want to acquire more possible solutions to increase the accuracy rate. So we modify this method to deliver the purpose. in the end, by means of emotion, age and gender, we recommend an advertisement which is high attention to customers, and record some information as a reference to achieve capability of Personal, consumerization and advertisement.

參考文獻


[6] Paul Viola and Michael J. Jones,“Robust Real-Time Face Detection”, International Journal of Computer Vision, Vol. 57, pp.137-154, May 2004.
[8] Peter N. Belhumeur, Joao~ P. Hespanha, and David J. Kriegman, "Eigenfaces vs. Fisherfaces: Recognition Using Class Specific Linear Projection", IEEE TRANSACTIONS ON PATTERN ANALYSIS AND MACHINE INTELLIGENCE, VOL. 19, NO. 7, JULY 1997
[10] Wikipedia - Linear discriminant analysis https://en.wikipedia.org/wiki/Linear_discriminant_analysis
[13] Labeled Faces in the Wild, http://vis-www.cs.umass.edu/lfw/
[16] Ahonen, T., Face Description with Local Binary Patterns: Application to Face Recognition, Dept. of Electr. & Inf. Eng., Oulu Univ. ; Hadid, A. ; Pietikainen, M., 2006

被引用紀錄


陳厚均(2016)。利用嵌入式平台實現居家人物觀看數位電視節目之行為分析〔碩士論文,國立中正大學〕。華藝線上圖書館。https://www.airitilibrary.com/Article/Detail?DocID=U0033-2110201614064853

延伸閱讀