Structured light vision measuring method based on three-dimensional surface contour of the k-means clustering color

Object of the present invention provides a method for measuring light visual surface structure of the object three-dimensional contour based on k-means clustering colors, in RGB space is determined for each linear projection color space, is calculated for each pixel to the color space of the linear Euclidean distance selecting the minimum value, corresponding to the color reference distance is closest to the current pixel color value. Then using the pixel points having the same reference numerals, the adaptive tuning parameters to fit a straight line of a new color space. Using the k-means clustering algorithm, iterative, until the color value of each pixel stabilized. After color classification to obtain the decoded sequence, and to determine the location of each matching point, generating a matrix of coordinates of the matching point. Using a triangulation method to obtain depth information, reconstruct a three-dimensional surface of the object. The invention improves the accuracy of color-coded, is robust and strong adaptive capacity to address the binocular stereo vision matching difficult.