CVL-DataBase

CVL 数据库是一个用于作家检索、作家识别和单词识别的公共数据库。该数据库包含 7 种不同的手写文本（1 种德语和 6 种英语文本）。共有 310 位作者参与了该数据集。其中27人写了7篇文章，283名作家写了5篇文章。对于每个文本，包含手写文本和打印文本样本的 rgb 彩色图像 (300 dpi) 以及裁剪版本（仅手写）可用。唯一的 id 标识作者，而每个单词的边界框存储在 XML 文件中。
CVL 数据库由从文学作品中选择的带有草书手写的德语和英语文本的图像组成。所有页面在右上角都有一个唯一的作者 ID 和文本编号（用破折号分隔），然后是打印的示例文本。文本位于两个水平分隔符之间。在印刷文本下方，要求个人使用直纹底纸书写文本，以防止文本行卷曲。布局遵循 IAM 数据库的样式。

The CVL Database is a public dataset for writer retrieval, writer identification and word recognition. It contains 7 distinct handwritten texts (1 German and 6 English texts). A total of 310 authors participated in this dataset. Among them, 27 authors wrote 7 texts, while 283 authors wrote 5 texts. For each text, both RGB color images (300 dpi) of handwritten and printed text samples, as well as cropped versions containing only the handwritten parts, are available. Each author is identified by a unique ID, and the bounding boxes of each word are stored in XML files.
The CVL Database consists of images of German and English cursive handwritten texts selected from literary works. All pages have a unique author ID and text number (separated by a dash) in the upper right corner, followed by the printed sample text. The text is placed between two horizontal separators. Below the printed text, participants were required to write the text on lined paper to prevent text lines from curling. The layout follows the style of the IAM Database.

搜集汇总

背景与挑战

背景概述

CVL-DataBase是一个用于作家检索、识别和单词识别的公共数据库，包含310位作者的德语和英语手写及打印文本样本，每个样本配有详细的元数据和边界框信息。数据集特别适用于手写文本识别和作家识别的研究。