The realization of identification method for DataMatrix code

Abstract

The minimum size of the DataMatrix code is 0.0002 square inches and this size is the smallest of the current one-dimensional and two-dimensional codes. So it is particularly suitable for printing on metal parts of circuit boards. However, compared with the general two-dimensional code, there are still many technical difficulties on metal surface DataMatrix code identification technology. The reflection on metal surface is not conducive to the DataMatrix binarization. The large spacing between the codes is not conducive to positioning and identification. There is a distorted or incomplete DataMatrix code on metal surface. In this paper, a DataMatrix code recognition method is proposed by digital image processing technology. Firstly, Binarization is achieved by improving the traditional Otsu algorithm. The binarized image is dilated to solve the problem of large code interval. Secondly, the DataMatrix code region is located. The Hough transform is used to detect the four vertices of the DataMatrix code. The vertex is used to correct the twisty DataMatrix code. Finally, according to the structural characteristics of the DataMatrix code, the incomplete code is completed. According to the encoding rules, the DataMatrix code is decoded. After the above series of steps, we can achieve the purpose of identifying DataMatrix code.