선택한 단어 수는 입니다.
최소 단어 이상 선택하여야 합니다.
최소 단어 이상 선택하여야 합니다.
선택한 단어 수는 30입니다.
최대 10 단어까지만 선택 가능합니다.
최대 10 단어까지만 선택 가능합니다.
제안 방법
- Apart from this, a combination algorithm of PCA and LDA was also proposed in this paper in order to carry out a recognition performance comparison rate between two algorithms and to distinguish which algorithm was suitable for certain specific system. The PCA + LDA algorithm face recognition achieved by the proposed estimator, based on 6 eigenvectors.
- And then, the 37 eigenfaces are computed into the normalized weight vectors in order to perform the recognition task. In the paper, the recognition task is carried out through two testing sets. Let the testing set from ICONL database to be 9s = [(Psi, (Ps2, …’, (Psm] and transformed it into its eigenface components or coefficients asws = U7C?((p5for S = .
- Basically, there are four basic threshold values, two mean values - ml and m2, and two variance values -vl and v 2. The recognition performance for each threshold and for the combination of those thresholds are simulated according to the face classification procedures with the idea to move the thresholds among crossing point of the two distributions (mean and variance). The classification outcomes are summarized in Table 1 and 2, respectively.
대상 데이터
- i) The training set. It consists 60 images of ICONL database of 12 subjects with 5 images of different expression each and are selected according to numerical ascending of 12 subjects with odd numeric. It is used to form a face space in which the recognition is performed.
- The database contains 140 images of 14 subjects - 2 females and 12 males of students and staff of engineering faculty with the age range of 24 to 55 years old. The photographs were manually taken by clicking on a mouse and were immediately stored inside the computer in RGB format with the resolution 1024 x 768 pixel in bitmap file.
이론/모형
- The face recognition using PCA algorithm was achieved by the proposed estimator, based on 37 eigenvectors. The simulation results showed the algorithm was satisfactory for recognizing an individual but was somewhat poor in classifying an individual within his or her face class.
- (ICONL) database. The preprocessed database is used to examine the adaptability of face recognition algorithm of the paper. The adaptability and the robustness of the algorithm are simulated by classifying the face images as a face or non-face and then as a known or unknown individual to the initialized training set.
- But, the attractive viewpoint of this concept lies in the fact that the 'face space' has the low dimensions and can make substantially their dimensions reduce without any loss of the image resolution. We can develope various efficient systems based on this key idea, the most notably by the Media Laboratory at MIT; see [2, 3, 4], The proposed face recognition approach in this paper is based on this key word, the Principal Component Analysis (PCA) algorithm.
※ AI-Helper는 부적절한 답변을 할 수 있습니다.