Research And Implementation Of Image Edge Detection Technology In Face Recognition

The development of computer technology provides a broad development space toimage processing technology. The research and application of image processingtechnology has been used in many disciplines and fields, it has been used more andmore in the field of security, such as face recognition technology, which has practicalsignificance already take an important part in the security verification, monitoring andmanagement.In the thesis, the face detection software is designed for security demonstrationproject for Yu Xiang Technology Co., Ltd., Harbin City, The project required the designof security management system software, and security management involves manytechnical, one on the need to implement a face detection function. The security systemsoftware is modular in design, operates independently of each part, and required to bereference by the main program directly. User requirements for face detection softwareshould running under the Windows operating system, and be able to build an interfacebetween the main program and camera, to give a clear result of the existence of humanface by detecting the image.Face detection method to achieve the functional requirements of the softwareapplications has mainly two elements constituted by the color matching and edgedetection, which is how to achieve accurate and effective edge detection, face detectionis the key. There are many algorithms for edge detection in the field of image processing,but so far, there are still many imperfections in relevant theory and algorithms.Therefore, optimized edge detection algorithm according to the specific applicationrequirements of the task of face detection is the necessary work to achieve the coresoftware algorithm, and most the most important work for completing the software.To complete the design of the face detection software under the Windows operatingsystem, this thesis studied the image processing technology, image edge detectiontechnique, wavelet transform theory and face detection technology; prepared for therealization of software algorithms and code article, this thesis first face of modeling andimage pre-processing method, based on the skin color the polymerizable of face ellipse model and image preprocessing, including the form of image binarization and imageprocessing, multiscale edge detection methods and wavelet denoising realize the imageedge detection, so as to achieve the functionality of the detected face Matlab simulationto test the effect of edge detection.To achieve full functionality of the software, including the interface with thehardware and call relationship with the security management system, this thesis studiedin software development platform, using the Windows API and C language, realized thefunction such as the video capture, face detection algorithm and the output of results.Complete software testing, the test results show that the stable operation of the software,and can effectively detect the existence of the human face information to achieve theexpected design requirements to provide a solution path, face recognition securitysystems.