摘要
近年来,基于深度学习的场景文字检测和识别研究已成为计算机视觉领域的一个研究热点。本文首先介绍了场景文字检测与识别所面临的挑战,其次从场景文字检测、场景文字识别和端到端文字识别三个任务分别综述了最新的研究工作,然后列出了该领域比较常用的大型公开数据集情况,最后总结和展望了最新的研究趋势。
In recent years, deep learning based scene text detection and recognition has become a research hot spot in the field of computer vision. The paper first introduces the challenges of scene text detection and recognition. Secondly, we review the latest research work from three tasks: scene text detection, scene text recognition and end-to-end text recognition. Then we list the opened big data sets commonly used in this field, and finally summarize and look forward to the latest research trends and focus.
作者
艾合麦提江·麦提托合提
艾斯卡尔·艾木都拉
阿布都萨拉木·达吾提
AHMATJAN Mattohti;ASKAR Hamdulla;ABDUSALAM Dawut(College of Information Science and Engineering,Xinjiang University,Urumqi 830046,China;School of Software,Xinjiang University,Urumqi 830046,China)
出处
《电视技术》
2019年第14期65-70,共6页
Video Engineering
基金
国家自然科学基金(61662076)
关键词
深度学习
场景文字
文字检测
文字识别
端到端识别
deep learning
scene text
text detection
text recognition
end-to-end text recognition