摘要
随着深度学习算法的兴起和普及,在人工智能领域取得了令人瞩目的进步。如今在AI图像识别领域更是发展迅速,通过AI图像识别可以将以前人工智能完成的任务转换为由计算机自动完成。设计实现了一个AI图像识别工具,首先对图片进行预处理,对图片进行校正和去噪,然后切割图片,进行识别和修正,从而对身份证、银行卡和驾驶证等进行识别,并自动提取其中的信息。其中,对文本的识别采用深度学习模型完成。对该AI图像识别工具进行测试,结果表明,该工具运行稳定,识别率在95%以上,符合最初的设计要求。
With the rise and popularity of deep learning algorithms,the field of artificial intelligence has made remarkable progress.Nowadays,the field of AI image recognition is developing rapidly.Through AI image recognition,the tasks completed by artificial intelligence before can be transformed to be completed automatically by computer.Design in this paper implements a AI image recognition tool,the tool for image preprocessing,first to correction and denoising of images,and cut the image,the final identification and correction,which can identify the id card,bank card,driving license,etc.,and automatically extract the information,including the recognition of the text is a deep learning model.Through the test of the AI image recognition tool,the test results show that the tool runs stably and the recognition rate is more than 95%.
作者
黄率
Huang Lyu(College of Information Engineering,Hunan University of Science and Engineering,Yongzhou Hunan 425199)
出处
《现代工业经济和信息化》
2023年第1期90-91,94,共3页
Modern Industrial Economy and Informationization