Data Matrix Two-dimensional Bar Code Pre-processing Method

With the development of information technology, the Bar Code has became one of the most economic and flexible information carriers. As a result, it plays an important part in the field of current industrial automation. In the semiconductor industry, refined management has also been further promoted. And each material bar is expected to be recorded and tracked in real time. Especially, the Two-Dimensional Bar Code which is viewed as an effective information carrier has been widely used. Hence, it plays a very vital role in the detection process of the Two-Dimentional Bar Code.As the influence of semiconductor industry processes, the Two-Dimensional Bar Code may be polluted or battered. At the same time, noise and blur will further lead to an increasing difficulty and reduce the accuracy of identification. Therefore, the paper will address a widely used bar code in the semiconductor industry which is named as Data Matrix to design a pre-processing algorithm for this kind of images and improve the quality of the images.In this paper, I analyze the symbol characteristics and coding structure of Data Matrix. And based on the traditional pre-processing method of the Two-Dimensional Bar Code images, I designe an improved pre-processing algorithm for Data Matrix images successfully. From the experimental results we can see that the proposed pre-processing algorithm demonstrates strong applicability and good robustness. According to the actual status of industrial production, I obtain an initial target region of interest from the original image, and denoise the target image of interest. At last, I use the knowledge of morphology to enhance the image by the the White Top Hat and the Dilate Transformation. Meanwhile, the experiments verify the feasibility and effectiveness of the method. In addition, I achieve the design work of software codes and hardware platform. Through simulating and testing in actual system, the proposed algorithm shows high efficiency and reliability.The value of this paper is to apply theory to reality, and I developed an effective solution to identify the Data Matrix in semiconductor industry. The proposed pro-processing algoriths are satisfy the requirements of the actual industrial production, but also save the cost and improve production efficiency. As a consequence, it has a better actual value.