FPGA-based Bar Code Image Edge Detection Algorithm And Its Hardware Design

In recent years,barcode identification system had been widely used in industrial production,logistics management and other fields,bringing great convenience to human social life.Meanwhile,higher and faster requirements had been put forward for the applicability,real-time performance and accuracy of barcode identification system.Bar code recognition equipment based on ARM and other single-architecture hardware platforms can not meet the high real-time requirements of industry and logistics.In order to solve the above problems,this paper studied the filtering and edge detection algorithm of barcode image and used FPGA to realize algorithm mapping.The hardware system of barcode recognition was designed,and the system was managed by ARM to effectively improve the real-time performance and applicability of the system.The main contents of this paper are as follows:(1)Research on barcode Image filtering algorithm based on FPGA.In order to solve the problem of fuzzy barcode edge contour caused by noise point diffusion in traditional median filtering algorithm,a switching mean-median filtering algorithm based on FPGA was proposed.Noise switch was used to identify noise points and median filtering was performed.Then,the threshold comparison module based on the mean value was used to identify the edge contour area of the line,and the non-edge contour pixel points were filtered to protect the edge information of barcode image lines.At the same time,the time sequence optimization of the median sorting process of the algorithm was carried out through the extreme value comparator,which saves hardware resources.(2)Research on barcode image edge detection algorithm based on FPGA.To solve the problem that classical sobel algorithm required high illumination environment and manual threshold setting was required when it performs edge detection of barcode images,a multi-orientation threshold self-adaptive sobel algorithm based on FPGA was proposed.Directional convolution template was added on the basis of the classical sobel algorithm to make the edge information detected by the algorithm more complete.Then,multiple threshold values of the image were collected in real time using the mean valuemodule to achieve threshold self-adaptation and reduce the requirements of the algorithm on the illumination environment.(3)Hardware design of barcode recognition system.Aiming at the real-time and applicability requirements of barcode identification system in the industrial field,the advantages and disadvantages of the existing embedded hardware architecture were compared and analyzed.Then combined with the system parameters and functional requirements,the barcode recognition system based on FPGA and ARM was designed and the peripheral circuit design was completed.Finally,the improved barcode image budget algorithm is verified on the hardware platform of barcode recognition system designed in this paper.The results show that compared with the traditional algorithm,the improved algorithm had better filtering effect on barcode image,and effectively protects the line edge information.The improved edge detection algorithm realized threshold adaptive,and the detected barcode contour feature information is more clear and complete.