Three-Dimensional Shape Reconstruction For Dynamic Object Based On Structure Light

Image recording equipment captures two dimensional information of the world with depth information missed. There are many areas need the three dimensional information of the object, just like industrial production, reverse engineering, medical diagnosis and so on. Three dimensional (3D) shape reconstruction techniques develop quickly with the needs of practical application. The color coded structure light method is becoming popular because of its advantages which are non-contact, high precision and easy to implement, It contains two kinds of encoding methods:timing sequence encoding method and spatial encoding method. The paper studied both of the encoding methods, and presented a novel spatial encoding method. At last we realized 3D shape reconstruction for dynamic object.The main content of the paper are as follows:Firstly, the paper analyses the history and current status of the 3D reconstruction technique, summarizes the advantages of the color coded structure light method by the comparison of the different methods, sets up the mathematical model of the reconstruction system according to the principle of the color coded structure light method.Secondly, the paper analyses the relationship of numbers between the color and the patterns in the timing sequence encoding method. Based on the analysis, a timing sequence encoding method is proposed. Then, the experiment result proves the method is practicable to reconstruct static object. Many effective image processing methods is used in the process.Thirdly, the paper focus on the 3D shape reconstruction for dynamic object by the spatial encoding method. The paper presents a novel spatial encoding method using permutation and combination method, the point is that this method uses less number of deferent colors and has high-accuracy in decoding. There are four stripes in each combination, different stripes in one combination may have the same color, so it can cut down the number of colors and reduce the difficulty of color discrimination. The experiment results demonstrate that the method can reconstruct the dynamic object.Finally, error analysis of the experiment results is given. The precision of the 3D reconstruction system is analyzed based on the experiment data.