| Summary: | 碩士 === 南台科技大學 === 電機工程系 === 97 === Recently, since more people pay much attention to the security of life and property, there are more applications of identification based on human features. For example, access control and management system, home security surveillance system, and robotic interactive recognition system are widely available.
The thesis proposes a human face recognition system based on the integration of principal component analysis (PCA) and linear discriminant analysis (LDA). It aims to find out the eigenvalues, eigenvectors, and eigenspace of human facial features using PCA firstly, and then obtain the data of facial weightings by projecting the eigenvalues to eigenspace of human face. The purposes of integrating LDA to the PCA based recognition scheme are not only to reduce the dimension of the images, but also to reduce the level of the image isolation in different categories by LDA to expend the distances between each central point of different categories. After these, one can determine the magnitude of Euclidean distance to make the recognition decision of human faces. These will accomplish the final facial recognition.
|
|---|
