Abstract
This invention provides one kind of characteristic points automatically identification method for three-dimensional space scanning data of human body in which important characteristic points of human body are identified on the basis of the principles by simulating human vision and touch without sticking marking points in advance and several calculating methods are developed in view of the property of each characteristic point and characteristic line including: outline analysis: projects the three-dimensional space data of human body to the two-dimensional plane and the positions of characteristic points are searched by utilizing the variations of curvature and depth; the minimal body surface circumference: according to its variation searches the thinnest visual part to define the characteristic point position; detection of the gray level value: transforms the color information of human body to the gray level value to simulate the darker visual feeling and finds the parts with obvious brightness variation and shadow; and human body contour line: simulates the most protruding or the most hollow situation of human touch feeling to find the highest and the lowest points. With this invention, the characteristic points of human body could be found quickly, further a lot of items of human body sizes could be stably and effectively obtained, and this invention could be helpful to provide the application and development of fields like customized clothes manufacturing, industrial design, human-factors engineering and so on.