39,993张互联网图像OCR数据【数据堂】

39,993张互联网图像OCR数据采集场景包括影视字幕、广告、手机截屏、漫画、表情包、海报、杂志封面等。语言分布为中文和英文（少量）。在标注方面，对互联网图像进行行级矩形框标注、行级内容转写（少量数据为列级矩形框标注、列级内容转写） 。本套互联网图像OCR数据可用于多种互联网图像OCR任务

This dataset contains 39,993 internet images for OCR tasks, with collection scenarios covering film subtitles, advertisements, mobile phone screenshots, comics, memes, posters, magazine covers and other scenarios. The languages involved are Chinese and a small amount of English. For annotation, most of the internet images are labeled with line-level bounding boxes and corresponding line-level content transcriptions; a small subset of the data uses column-level bounding boxes and column-level content transcriptions instead. This set of internet image OCR data can be applied to various internet image OCR-related tasks.

搜集汇总

背景与挑战

背景概述

该数据集包含39,993张互联网图像，涵盖影视字幕、广告、手机截屏等多种场景，主要用于OCR任务。数据以中文为主、英文为辅，采用行级矩形框标注和内容转写，标注精度不低于97%。此为商业数据，仅面向企业合作购买。