A new method is proposed aiming at decoding De Bruijn color encoded structured light images with high precision.This method is composed of the central color stripe identification based on gradient analysis

the color classification based on clustering analysis and the feature point matching based on double-gradient and multi-pass dynamic programming.The method can be applied to avoid color interference effectively in the encoded structured light images containing parallel pixel lines with different colors.Moreover

by making full use of local neighborhood information

the method can match every feature point.Experimental results show that the proposed method effectively improves the decoding accuracy without assuming that the surface of the target object is monotonous or restricting its continuity

and also exhibits good robustness.