|
In recent years, along with the arrival of the information age, biometric identification technology in the world, the development of faster and faster, a variety of feature recognition technology has entered our lives, and has been successful to a certain extent. For example: face recognition, fingerprint recognition, iris recognition, voice recognition, facial expression recognition, gait recognition, signature recognition, and so on. But on the whole each biometric and there are some limitations. If we can find a less camouflage, and is not subject to changes in the external conditions change biometric based biometric identification technology will be a big breakthrough. The skull is the most complex and most important part of the human skeleton, it contains a lot of the people, and the skull characteristics vary from person to person, this series of conditions skull-based feature recognition technology has a great practical skills. However, due to the two-dimensional skull characteristics contains less information, and the shape of the unstable characteristics is obvious. In this paper, based on the identity of the three-dimensional skull feature recognition. Its main tasks are the following aspects: First, a lot of references, discusses the purpose and significance of the the skull feature recognition research and analysis of relevant aspects of the research results at home and abroad and the latest progress; Second, the skull image preprocessing, This is one of the key steps of image noise removal, image interpolation and image segmentation. Lay a good foundation for the back of the three-dimensional reconstruction; then describes two three-dimensional reconstruction algorithm, surface rendering, volume rendering, analysis of the advantages and disadvantages of both, in view of the purpose of this paper is to model surface coordinate information, so we used surface rendering surface reconstruction algorithm for two-dimensional skull sections. Since everyone has a different three-dimensional shape, so it in the space in the data coordinates are not the same, the use of Visual C 3.0 software programming to obtain three-dimensional coordinate values ??of the image of the skull, and the data is stored; again, since the amount of data of the entire space large, then we will have three-dimensional coordinate data compression, this article uses a wavelet transform to extract inside information; Finally, select an appropriate classification and identification method to identify the characteristics of the skull. This paper focuses on the three-dimensional coordinates of the skull, after several test results show that this feature of the three-dimensional coordinate information of the skull has played a good role in the skull recognition process.
|