The Research And Implementation Of Number Recognition Based On FPGA

With the rapid development of computer technology, digital image processing technology and microelectronic technology, the FPGA as the core of the image processing system has been widely used in various fields.In this paper, FPGA is used as a platform for the research and implementation of the digital recognition of printed body in the visual calculation. In this paper, the background of the subject and the research of visual computing are briefly introduced. And based on it, a variety of digital image recognition algorithms and image preprocessing algorithms are discussed. According to the simple background of digital image recognition, we use the median filtering algorithm, the two-peaks method, the segmentation algorithm based on Projection and the recognition algorithm based on the digital features to implement the system. And we have improved the implementation of these algorithms, and make it more suitable for FPGA implementation.According to the system requirements, the overall structure of the FPGA chip is described in detail in this paper. We have designed a digital image acquisition module, dual RAM memory module, image preprocessing module, digital recognition module, PLL clock phase lock loop and so on in the FPGA.Based on the IP module of FPGA, this paper introduces the design and implementation of the digital recognition algorithm in detail. In order to reduce the consumption of SRAM resource and invalid data, and to improve the efficiency of system operation, the system use a variety of image preprocessing algorithms to extract useful information while filtering out useless information. Finally, we use the method of edge detection to identify the digital features of the image. We use the idea of parallel to design the system. This system uses double RAM to store data under Ping-Pong operation. And we design a parallel image processing system which is composed of two independently operating system. It can be used to handle the data from two pairs of RAM, and make full use of the parallel characteristics of FPGA, and improve the operating speed of the system.In this paper, we use Quartusii, MATLAB, Modesim and other tools to implement each module of the system in altera FPGA. We have carried out a series of simulation and comprehensive verification for the system. Simulation and test results show that the digital recognition system designed by this paper can identify one or more numbers, and the accuracy rate is also achieved.