自然场景OCR描述数据

20,000张自然场景OCR描述数据，包含亚洲语系与欧洲语系，总计14种语言，采集环境包括商店牌匾、站牌、海报、路标等多种场景，含有多种拍摄角度。描述语言为英语，主要描述文字排列方式、文字内容，颜色等信息。

This dataset includes 20,000 natural scene OCR description samples, covering Asian and European language families with a total of 14 languages. The data was collected from diverse scenarios such as store signs, bus stop signs, posters, road signs and more, with multiple shooting angles. The annotation language is English, which mainly describes details including text arrangement, text content and color.

搜集汇总

背景与挑战

背景概述

该数据集包含20,000张自然场景图像，涵盖14种亚洲和欧洲语系，采集自商店牌匾、站牌等多种环境。图像描述使用英语，重点描述文字排列、内容和颜色等信息，但仅限企业合作购买使用。