摘要
传统应用架构下,识别的精度会发生不同的变化,很难对其进行控制,时常出现精度不高、耦合范围无法确定等问题。因此,本文基于现场可编程逻辑门阵列(Field Programmable Gate Array,FPGA)架构的手写数字智能识别系统进行设计。通过对Web服务器识别处理器以及重叠识别电源电路的设计,完成硬件设计;对FPGA识别分析目标模块以及FPGA架构下识别数据库进行设计,完成软件的设计。最终的测试结果表明:对比于固定的识别标准,系统所测试的实际识别标准均达到了预期的目标,FPGA手写数字智能系统实际的识别效果更佳,误差也相对较小,具有潜在应用意义。
Under the traditional application framework, the recognition accuracy will change differently, it is difficult to control it, and problems such as low accuracy and uncertain coupling range often occur. Therefore, the handwritten digit intelligent recognition system based on FPGA architecture is designed. The hardware design is completed by identifying the processor of the Web server and the design of the overlapping identification power circuit;then the FPGA identification and analysis of the target module and the identification database under the FPGA architecture complete the software design. The final test results show that: compared with the fixed recognition standards, the actual recognition standards tested by the system have reached the expected goals, indicating that the actual recognition effect of the FPGA handwritten digital intelligent system is better, the error is relatively small, and it has potential application significance.
作者
李素娟
LI Sujuan(Hebi Polytechnic,Henan Hebi 458030,China)
出处
《信息与电脑》
2021年第23期95-97,共3页
Information & Computer
基金
中国高校产学研创新基金—异构智能计算项目“基于HLS的手写体数字识别”(项目编号:2020HYA01001)。
关键词
手写数字技术
智能识别系统
FPGA架构
handwritten digital technology
intelligent identification system
FPGA architecture