OCR Datasets

This dataset contains handwritten text images collected from 100 individuals, including 50 Japanese, 49 Koreans and 1 Afghan. For different subjects, the corpus are different. The data diversity includes multiple cellphone models and different corpus. This dataset can be used for tasks such as handwriting OCR models, handwritten text recognition systems, and multilingual OCR pipelines

This dataset contains 1,586,458 document data sets covering a wide range of Chinese-language materials, including Chinese textbooks, e-books, and teaching reference materials. The annotation files provide both OCR annotations and structured document analysis. The dataset is suitable for document OCR, document understanding tasks.

This dataset includes 9,574 handwriting images across 8 languages, including English, Spanish Portuguese and more. The data diversity includes multiple collecting scenes, different text carriers and different photographic angles(looking up, eye-level, looking down). In terms of annotation, each text line is annotated with quadrilateral polygons and transcription. The dataset can be used for training and evaluating OCR models, handwriting recognition systems, and multilingual text extraction tasks in AI and computer vision.

426,687 high-resolution images featuring multilingual Optical Character Recognition (OCR) data across both natural scenes and various document types. This dataset spans 20 languages, including Traditional Chinese, Simplified Chinese, Japanese, Korean, Thai, Vietnamese, Indonesian, Malay, Polish, and more. The data covers a wide range of real-world conditions—natural scenes, printed documents, handwritten notes, signs, and posters—captured from multiple countries and environments, with varied backgrounds, lighting conditions, and camera angles. All images are annotated for OCR tasks, making this dataset highly suitable for training deep learning models for text detection, recognition, and layout analysis in multi-language scenarios. The dataset complies with global data protection standards (GDPR, CCPA, PIPL), and is validated by leading AI enterprises for commercial and research applications.

This dataset contains 5,162 handwriting images from 262 individuals, covering Traditional Chinese characters used in Taiwan. Each text in the data were annotated with quadrilateral bounding boxes. The handwriting ocr data can be used for training and evaluating OCR models, Traditional Chinese character recognition systems, and AI-based handwriting applications. The accuracy of line-level annotation and transcription is >= 97%.

5,000 Images-4 Categories Cards OCR Data, applicable for tasks such as character recognition.

102,870 Sets - Chinese Instruction Manual OCR&Parsing Data, including Drug instructions, Installation instructions, Operating Instructions, Product manual, User manual, the annotation files contain OCR annotations and structured parsing.

14,980 Images PPT OCR Data of 8 Languages. This dataset includes 8 languages, multiple scenes, different photographic angles, different photographic distances, different light conditions. For annotation, line-level quadrilateral bounding box annotation and transcription for the texts were annotated in the data. The dataset can be used for tasks such as OCR of multi-language.

15,148 Sets - Japanese and Korean Menu OCR Data, including 9,701 Japanese menus and 5,447 Korean menus, annotation content includes row level quadrilateral box annotation, row level content transcription (a small portion of data is column level quadrilateral box annotation, column level content transcription)