摘要
文本检测技术在社会中有着广泛的应用,随着深度学习的加入,文本检测技术得到了进一步的提升。近年来基于深度学习的检测算法逐渐增多,针对场景文本检测的各种问题提出了相应的解决方法,提升了场景文本检测算法的性能。本文对这些算法进行了归纳、分析和总结,将这些算法大致分为基于回归和基于分割两种类型,并对其性能进行了对比,最后基于这些算法的研究内容为文本检测领域未来的发展提出了新的研究方向。
Text detection technology has a wide range of applications in society,and with the integration of deep learning,it has been further enhanced.In recent years,the number of detection algorithms based on deep learning has gradually increased,and corresponding solutions have been proposed for various problems in scene text detection,improving the performance of these algorithms.This paper summarizes,analyzes,and concludes these algorithms,categorizing them into two main types:regressionbased and segmentation-based.Their performances are compared,and based on the research on these algorithms,new research directions are proposed for the future development of the text detection field.
作者
张静
孙巧榆
刘珍兵
ZHANG Jing;SUN Qiaoyu;LIU Zhenbing(School of Electronic Engineering,Jiangsu Ocean University,Lianyungang Jiangsu 222005,China)
出处
《智能计算机与应用》
2024年第2期48-54,共7页
Intelligent Computer and Applications
