Image processing of seam tracking system with laser vision

The method of real time image processing in the laser vision sensing seam tracking system was studied.In the course of pretreatment, the contrast of image was intensified by the image intensify treatment; the noise in the image was removed by medium filtering treatment;and the aim image was obtained from the back ground image by the binariantion treatment.Then in the image posttreatment, for the key technology in laser stripe extraction and character point detection, some feasible method was proposed.The single and consecutive laser stripe was obtained by the method of inside stalk transformation.The convenient and reliable characteristic pointswere detected by analyzing slop.The method proved that the character point canbe detectedexactly and rapidly, and the demand of real time seam tracking system can be meeted by this method.