中英文ocr数据集

该数据集包含200万张用于文字识别的图像样本，其中英文和中文各100万张。数据集来源：英文样本：DonkeySmall/OCR-English-Printed-12中文样本：来自https://github.com/YCG09/chinese_ocr，只包含Chinese_dataset； 100万训练数据

This dataset contains 2 million image samples for optical character recognition (OCR), including 1 million samples for English and 1 million samples for Chinese, respectively. Regarding data sources: the English samples are sourced from DonkeySmall/OCR-English-Printed-12; the Chinese samples are exclusively taken from the Chinese_dataset subset of the repository hosted at https://github.com/YCG09/chinese_ocr. This dataset consists of 1 million training samples.

搜集汇总

背景与挑战

背景概述

该数据集包含200万个高质量印刷文本识别图像样本，其中中英文各100万。它适用于计算机视觉任务，如文本检测和识别，数据来源于两个开源项目并经过精心预处理。