| Nowadays, Power plants provide continued impetus to the development of social. The safety of power plants guarantees the rapid economic growth of our country. In large power plants, various production workshops are the focus of security. The establishment of a perfect, reliable access control system can provide help for production, intelligent office and automated management. Biometric has better reliability, security and availability with its innate, unique and constant characteristics. Comparing to other biometrics, face recognition technology has the advantage of non-intrusive, easy to operate and low acquisition cost, so it has been widely used in biometrics market at home and abroad. Face recognition access control system can closely integrate with the original technology in traditional areas of security, so it can greatly enhance the level of intelligence of the original system and promote the technological upgrading of traditional industries. In this paper, the goal is to design and realize the access control and attendance management system in power plant, the main research work and achievements are as follows:Firstly the overall design of the access control and attendance management system based on face recognition is given in this paper. In order to meet the system independence and scalability requirements while reducing the complexity of the system, the modular design is made and the performance indicator is given in this paper.In order to meet the system requirements for real-time, the AdaBoost algorithm is used in this paper. During the training of the classifier, the Haar-like feature and the Local Binary Pattern(LBP) feature are used separately. It is found that the algorithm combined with the LBP feature has a distinct advantage on speed. Considering the demand for computer performance is too high for the work of real-time face detection, and it is hard to maintain high frame rate detection, the face tracking technology is used in the paper. Additionally, the function of detecting moving objects is added to the module in order to reduce the unnecessary computational overhead. Finally, this module meets the needs of the system performance requirements, which is proved by experiments. The face recognition module is achieved in this paper. Firstly, the locations of eyes are used for geometric normalization of the image. To reduce the amount of computation, sampling is used to process the image. Considering the Gabor Wavelet is not sensitive to changes in brightness and facial expression, convolution operation is made with Gabor filters for face images, and the results are set as the feature vectors of the images. During the process of matching, the Principal Component Analysis (PCA) algorithm is used to reduce the dimensionality. Then the Linear Discriminant Analysis (LDA) algorithm is used to classify the face images of each group, and to make the second dimension reduction mapping of the eigenvectors. Finally, the Euclidean distance between the feature vectors is calculated to find the best match face images. The experimental results show that this module is able to meet the requirements of the computation speed and accuracy when the bias between the orientation of the user’s face and the forward direction is small.And then the access control module is realized. After the user’s identity is identified, the open signal is sent to the SCM through the serial port. Then the relay is switched off and the electronic door lock is opened. The SCM sends drive signals to the stepper motor to open the door. After the door is completely open in6seconds, the SCM controls the closure of the door and the lock.In the end the access control and attendance management system in power plant is achieved and tested in this paper. Firstly the database table about attendance information is designed, then the facial feature library is established and the complete system is built. The experimental results show that the system achieves the performance indicators. |
PDF Full Text Request