71,535 Images English OCR Data in Natural Scenes
- 71,535 images
- English OCR
- Natural Scenes
Datatang has passed the certification of ISO27001 Information Security Management System and ISO9001 Quality Management System.
Data Introduction
71,535 Images English OCR Data in Natural Scenes. The collecting scenes of this dataset are the real scenes in Britain and the United States. The data diversity includes multiple scenes, multiple photographic angles and multiple light conditions. For annotation, line-level & word-leve & character-level rectangular bounding box or quadrilateral bounding box annotation were adopted, the text transcription was also adopted. The dataset can be used for English OCR tasks in natural scenes.
Data Specification
- Data size
- 71,535 images, each image has 1-200 words
- Collecting environment
- onsite collection in Britain and the United States, including shop plaque, poster, road sign, reminder, warning, packing instruction, menu, building sign, etc.
- Data diversity
- including multiple scenes, multiple photographic angles, multiple light conditions
- Device
- cellphone, camera, tablet
- Photographic angle
- looking up angle, looking down angle, eye-level angle
- Data format
- the image data format is .jpg, the annotation file format is .json
- Annotation content
- line-level & word-level & character-level rectangular bounding box or quadrilateral bounding box annotation; transcription for the texts
- Accuracy
- the accuracy of bounding boxes annotation is not less than 95%; the texts transcription accuracy is not less than 95%
