Reserch On DataMatrix Code Recognition Algorithm In Complex Background

In the industrial field,the use of two-dimensional code of industrial products and parts to identify,to achieve the production of products and parts tracking,assembly management,life cycle maintenance has become the industry standard for the automation industry.Which Data Matrix(DM)two-dimensional code because of its excellent data compression capabilities and strong error correction capability by the industry and the logistics industry of all ages.Unlike the packaging of food,pharmaceuticals and other consumer products,the application environment of industrial two-dimensional codes is usually harsh,and the identification of two-dimensional codes is usually accompanied by problems such as noise,exposure,wear and contamination.In particular,the identification of DPM(Direct Part Marking)for direct components indicates that conventional two-dimensional code recognition methods can not meet their needs.Therefore,the DM code extraction algorithm for complex background is of great significance and urgent market demand.Based on the above research background and needs,this paper analyzes the technical background and application direction of the industrial two-dimensional code,and makes a deep research on the DM code on industrial products and components.A new method of DM code based on angle and edge information is proposed.Adaptive lattice division of the DM code sampling method to achieve the complex background of the DM code quickly and accurately identify.The main contents of this paper include the following aspects:1.Based on the two-dimensional code feature,a method of precise positioning of DM two-dimensional code based on corner point density and edge information is proposed,which includes four parts: rapid positioning and prioritization of DM code candidate region based on corner distribution;DM code candidate area outside the contour extraction,reduce the non-interested edge on the positioning of interference;improved Hough transform coarse positioning L side,fast linear voting;iterative weighted least squares linear fitting,precise positioning L side.Compared with the traditional Hough transform,the "L" shape method improves the accuracy,speed and robustness of the DM region in the complex background.2.By analyzing the characteristics of DM code,an adaptive sampling grid partitioning algorithm based on image gradient projection is proposed.By calculating the gradient of the DM image and plotting it in the vertical direction,the two-dimensional image is mapped into a one-dimensional signal,and the peak-valley extraction is carried out.The trough is the sampling line position.Finally,the anomaly sampling line is corrected according to the peak-valley distribution,so as to realize the adaptive sampling grid division of DM code.The experimental results show that the proposed method can effectively adapt the pollmed or worn DM codes to improve the robustness of DM codes in harsh environments.3.Based on the embedded DSP TMS320DM648 to build DM identification test system to verify the performance of the proposed method.The experimental results show that the proposed method has good recognition ability for DPM codes with different processing technologies of metal materials,and can adapt to the fast and accurate reading of industrial complex environment such as uneven illumination,low illumination and defacement.