Detecting Human Face with Ridge-Valley-Normal Model

Detecting human face has been a popular but complex research topic which has successfully attracted the special attention of academic scholars and experts due to its widespread applications in practice. However, several existing methods fail to work with large -slanting angles of faces. To o...

全面介紹

Saved in:
書目詳細資料
主要作者: Lam Thanh Hien, Do Nang Toan, Ha Manh Toan
格式: 圖書 Book chapter Dataset
出版: ĐHQGHN 2016
主題:
在線閱讀:http://repository.vnu.edu.vn/handle/VNU_123/10939
標簽: 添加標簽
沒有標簽, 成為第一個標記此記錄!
實物特徵
總結:Detecting human face has been a popular but complex research topic which has successfully attracted the special attention of academic scholars and experts due to its widespread applications in practice. However, several existing methods fail to work with large -slanting angles of faces. To overcome such drawback, this study further develops the application of the ridge & valley characteristics on human face and incorporates them with the well- kno wn facial normal model to effectively detect human face. Our proposed algorithm consists of two consecutive phases namely “learning phase” and “detecting phase”. It is found effective in the experimental test with a data set with numerous images collected from different sources. Moreover, the novel searching procedure not only results in a new set of faces but also displays an examined face corresponding to which face model, from which the actual direction of human face detected can be easily calculated.