The Implementation Of Chinese Character Recognition Based On SOPC

Chinese character recognition has always been one of the research hotspots in the field of image pattern recognition,and it is closely related to our daily life and work.It is often used in the reading and translation of books and documents,the sorting of statistical reports,and the sorting of express delivery and mail,It is often used in the reading and translation of books and documents,the sorting of statistical statements,the sorting of express delivery and mail,and the checking of tickets and documents,etc.In view of the current limitations of the PC-based Chinese character recognition technology in the current market in terms of power consumption,volume,and portability,and Considering providing technical support for applications such as office automation,mail sorting,and blind readers,etc.This paper takes the printed Chinese character document as the research object,studies and designs the Chinese character recognition system based on SOPC.Based on the principle of Chinese character recognition,this paper studies the preprocessing,feature extraction and matching recognition algorithms in the recognition process.Finally,the system uses a combined feature recognition algorithm that consists of multiple features and multi-level recognition.After the algorithm was selected,it was first run on the PC to detect its actual effect.After the normal operation,the algorithm was used as the core to build the SOPC system.The system is divided into two parts: hardware design and software design in the design process.The hardware part adopts the method of adding off-the-shelf and custom IP cores in accordance with the modular structure,and is generally divided into three modules of image input,processing and display,and uses the Verilog HDL language to complete the construction of the hardware system.The software part completes the initialization and driving of SOPC system peripherals,and the migration and integration of algorithms.Taking into account the realization of the above functions,this paper uses Altera's DE2-115 test board and the Terasic company's MTLC suite at the time of design,with Nios II soft processor as the core,with Quartus II and Nios II Software Build Tools for Eclipse as development tools.The test proved that the system architecture proposed in this paper is effective and feasible,the algorithm is transplanted successfully,and the hardware runs well.The design of Chinese character recognition system based on SOPC is realized.