Three-dimensional Footprint Of The Weight Of Surface Segmentation And Description Of Morphological Characteristics,

Recently, with the high-speed development of information technology and the computer network technology, the identity identification based on the human body's biological characteristic which is named biometrics has been the research focus in the field of information technology. The three-dimensional footprint which is left when the human body stands or walks has special retain with people's height, posture, gender and age. This relation has been confirmed and applied to long-term practice successfully in criminal cases. As a new research direction in biometrics, there are many elementary theories and the algorithm research topics in footprint biological characteristic recognition.This paper studies the biological characteristic recognition technology of the three-dimensional footprint. The data processing and research of feature extraction by three-dimensional surface data appearing in three-dimensional footprint, is not merely a subject for research with the important academic meaning in range image processing and pattern recognition, there are obvious application prospects too in the fields of criminalities, personnel security etc..Aimed at the differential geometry relations of three-dimensional curved surface in three-dimensional footprint, we research the method of three-dimensional footprint pressure surface segmentation by range image processing and characteristic parameters extraction on pressure surface by taking the knowledge of footprint identity identification into account. The research effect and main achievement made include:First, we establish the database of footprint image and the human body's physiological characteristic.Establishing large sample database of footprint image and the human body's physiological characteristic database is the basic issue of enriching footprint identification theory and researching biometrics technology. There is no such a footprint database both at home and abroad at present. This database will be used in footprint identification algorithm design, and offer the test data for domestic and international corresponding research institution simultaneously.Second, we use differential geometry principle to study footprint range image segmentation, and have realized the segmentation and extraction of three-dimensional footprint pressure surface.Aimed at the feature of footprint range image, the paper has defined the concept of the form index and proposed the segmentation method based on the mean curvature and Gaussiancurvature. But because of containing serious noise in footprint range image, we extract pressure surface area which is interested by combining footprint experience knowledge with man-machine interactive action, and the result is better.Third, we propose a group of characteristic parameters which describe the human body's biological characteristic by taking the knowledge of footprint identity identification into account.Feature description is the foundation of the computer auto-recognition, it can influence the result directly. The paper has proposed a kind of description method based on the global and regional characteristic, and has done the initial analysis. The global feature can reflect that the footprint leaves over the distribution form of people's foot skeleton very well; the centroid polygon can describe the position relation between five toes better. These features can reflect human's characteristic objectively.