The Research Of Face Detection And Recognition System Base On Labview

In the modern society, identity authentication can be found everywhere, and isgrowing with each passing day. While the traditional identity authentication methodcould never be able to satisfy the needs of the people, biological features which arehuman inherent attributes become the most ideal basis of identity verification.Compared with fingerprint, voice and other biological feature recognition, facerecognition has the characteristics of directness, friendliness, convenience, and is easyto be accepted by users.In this paper, the face detection and recognition system based on LabVIEW isconstructed by face recognition technology. The system is divided into four parts:image acquisition, face detection, image preprocessing, feature extraction andrecognition.Face detection is the first step for face recognition. Due to the wrong detection ofnon-face regions with skin characteristics caused by simply using skin detection, theface detection algorithm based on skin color model and template matching is proposedin this paper. In YCbCr color space, skin segmentation is realized by use of theGaussian probability skin model. And using a template matching method for faceverification, the location of human face is ultimately determined in the image. Theexperimental results show that, the algorithm can improve the efficiency of detection,and provide accurate information for subsequent identification.Image preprocessing is a key step in face recognition. In order to ensure theconsistency of face size and face image quality, the procedures of face imagepreprocessing are mainly included as follows: color image to grayscale image conversion, image enhancement, image normalization.Face feature extraction and face recognition are two important problems of facerecognition. Due to the disadvantages of extract features, we deeply investigated PCAand LDA face recognition algorithm, and presented the PCA and LDA combinedfeature extraction method with the minimum distance classifier for face recognition inthis paper. A large number of experiments show that the method is effective andfeasible, and improves the face recognition rate to some extent.Finally, the hardware and software platforms of this face recognition system wereintroduced. And the framework of the system was proposed, with refining functions ofeach module. Combined with MATLAB7.0, face detection and recognition system isrealized by using LabVIEW2009SP1and IMAQ Vision9image processing softwarepackage. The system has the advantages of simple operation, friendly interface, andhigh reliability.