摘要
近年来,卷积神经网络在图像的分类识别领域取得成功,并逐渐在许多嵌入式终端设备上得到应用.在Linux环境下采用QT开发框架,设计实现了一个基于卷积神经网络的手写体数字识别系统.该系统可以很方便的跨平台移植于各类嵌入式设备上.测试结果表明,该系统对手写体识别具有良好的识别效果.
In recent years, Convolutional Neural Network （CNN） has been successfully used in the field of imageclassification and recognition. In this paper, based on the QT development framework, we designed andimplemented a handwritten numeral recognition system based on CNN in Linux environment. The proposed systemcan be easily transplanted on various embedded devices. The test results show that the system has good recognitioneffect.
出处
《微电子学与计算机》
CSCD
北大核心
2018年第2期71-74,共4页
Microelectronics & Computer
基金
国家自然科学基金资助项目(61473009)
北京市自然科学基金资助项目(4174086)
