[{"@type":"PropertyValue","name":"Data size","value":"57,645 images, 528,553 bounding boxes"},{"@type":"PropertyValue","name":"Collecting environment","value":"including street scenes, plaques, billboards, posters, decorations, art lettering, magazine covers etc."},{"@type":"PropertyValue","name":"Data diversity","value":"multiple scenes, multiple fonts"},{"@type":"PropertyValue","name":"Language distribution","value":"Chinese, English (a few)"},{"@type":"PropertyValue","name":"Bounding box direction distribution","value":"324,399 vertical bounding boxes, 204,154 non-vertical bounding boxes"},{"@type":"PropertyValue","name":"Bounding box shape distribution","value":"34,936 rectangular bounding boxes, 220,716 polygonal bounding boxes, 272,901 parallelogram bounding boxes"},{"@type":"PropertyValue","name":"Data format","value":"the image data format is .jpg, the annotation file format is .json"},{"@type":"PropertyValue","name":"Annotation content","value":"vertical -level rectangular bounding box (polygonal bounding box, parallelogram bounding box) annotation and transcription for the texts; non-vertical rectangular bounding box (polygonal bounding box, parallelogram bounding box) annotation and transcription for the texts"},{"@type":"PropertyValue","name":"Accuracy","value":"The error bound of each vertex of a bounding box is within 3 pixels, which is a qualified annotation, the accuracy of bounding boxes is not less than 97%; The texts transcription accuracy is not less than 97%."}]
{"id":1226,"datatype":"1","titleimg":"https://res.datatang.com/asset/productNew/APY230117001.png?Expires=2007353720&OSSAccessKeyId=LTAI5tQwXnJZbubgVfVa1ep9&Signature=2y7kVqALs6RYrapJ0h7QUaon1ck%3D","type1":"147","type1str":null,"type2":"150","type2str":null,"dataname":"Vertical Text OCR Dataset with 57,645 Images and Polygon Annotations","datazy":[{"title":"Data size","desc":"Data size","content":"57,645 images, 528,553 bounding boxes"},{"title":"Collecting environment","desc":"Collecting environment","content":"including street scenes, plaques, billboards, posters, decorations, art lettering, magazine covers etc."},{"title":"Data diversity","desc":"Data diversity","content":"multiple scenes, multiple fonts"},{"title":"Language distribution","desc":"Language distribution","content":"Chinese, English (a few)"},{"title":"Bounding box direction distribution","desc":"Bounding box direction distribution","content":"324,399 vertical bounding boxes, 204,154 non-vertical bounding boxes"},{"title":"Bounding box shape distribution","desc":"Bounding box shape distribution","content":"34,936 rectangular bounding boxes, 220,716 polygonal bounding boxes, 272,901 parallelogram bounding boxes"},{"title":"Data format","desc":"Data format","content":"the image data format is .jpg, the annotation file format is .json"},{"title":"Annotation content","desc":"Annotation content","content":"vertical -level rectangular bounding box (polygonal bounding box, parallelogram bounding box) annotation and transcription for the texts; non-vertical rectangular bounding box (polygonal bounding box, parallelogram bounding box) annotation and transcription for the texts"},{"title":"Accuracy","desc":"Accuracy","content":"The error bound of each vertex of a bounding box is within 3 pixels, which is a qualified annotation, the accuracy of bounding boxes is not less than 97%; The texts transcription accuracy is not less than 97%."}],"datatag":"OCR,Multiple scenes,Multiple fonts","technologydoc":null,"downurl":null,"datainfo":null,"standard":null,"dataylurl":null,"flag":null,"publishtime":null,"createby":null,"createtime":null,"ext1":null,"samplestoreloc":null,"hosturl":null,"datasize":null,"industryPlan":null,"keyInformation":"","samplePresentation":[{"name":"/data/apps/damp/temp/ziptemp/APY230117001_demo1695809214287/APY230117001_demo/054889_demo.jpg","url":"https://bj-oss-datatang-03.oss-cn-beijing.aliyuncs.com/filesInfoUpload/data/apps/damp/temp/ziptemp/APY230117001_demo1695809214287/APY230117001_demo/054889_demo.jpg?Expires=4102329599&OSSAccessKeyId=LTAI8NWs2pDolLNH&Signature=tDNg2Jmp1JEEZiTMQB2hIFr17SI%3D","intro":"","size":0,"progress":100,"type":"jpg"},{"name":"/data/apps/damp/temp/ziptemp/APY230117001_demo1695809214287/APY230117001_demo/056472_demo.jpg","url":"https://bj-oss-datatang-03.oss-cn-beijing.aliyuncs.com/filesInfoUpload/data/apps/damp/temp/ziptemp/APY230117001_demo1695809214287/APY230117001_demo/056472_demo.jpg?Expires=4102329599&OSSAccessKeyId=LTAI8NWs2pDolLNH&Signature=sGLSMlt8%2BocGKzLN0YYIgRrTg78%3D","intro":"","size":0,"progress":100,"type":"jpg"},{"name":"/data/apps/damp/temp/ziptemp/APY230117001_demo1695809214287/APY230117001_demo/054886_demo.jpg","url":"https://bj-oss-datatang-03.oss-cn-beijing.aliyuncs.com/filesInfoUpload/data/apps/damp/temp/ziptemp/APY230117001_demo1695809214287/APY230117001_demo/054886_demo.jpg?Expires=4102329599&OSSAccessKeyId=LTAI8NWs2pDolLNH&Signature=2t3vBygA4k0HDuTQq9h%2ByOBmO8E%3D","intro":"","size":0,"progress":100,"type":"jpg"}],"officialSummary":"This dataset contains 57,645 images collected from diverse real-world text scenes, including street scenes, shop signs, billboards, posters, decorative text, art lettering, and magazine covers. The dataset includes Chinese text and a small amount of English text. Each image is annotated with text localization information and transcription labels, including polygon and quadrilateral bounding boxes for vertical and non-vertical text regions. This dataset is suitable for tasks such as scene text OCR tasks and multi-oriented text recognition.","dataexampl":null,"datakeyword":["ocr dataset","ocr training data","scene text dataset","text recognition dataset","text detection dataset"],"isDelete":null,"ids":null,"idsList":null,"datasetCode":null,"productStatus":null,"tagTypeEn":"Data Type,Language","tagTypeZh":null,"website":null,"samplePresentationList":null,"datazyList":null,"keyInformationList":null,"dataexamplList":null,"bgimg":null,"datazyScriptList":null,"datakeywordListString":null,"sourceShowPage":"ocr","dataShowType":"[{\"code\":\"0\",\"language\":\"ZH\"},{\"code\":\"1\",\"language\":\"ZH\"},{\"code\":\"2\",\"language\":\"EN,JP,PT,DE,KO,FR,ES\"},{\"code\":\"3\",\"language\":\"EN\"},{\"code\":\"4\",\"language\":\"JP\"}]","productNameEn":"57,645 Images - Vertical OCR Data in Text Scenes","BGimg":"","voiceBg":["/shujutang/static/image/comm/audio_bg.webp","/shujutang/static/image/comm/audio_bg2.webp","/shujutang/static/image/comm/audio_bg3.webp","/shujutang/static/image/comm/audio_bg4.webp","/shujutang/static/image/comm/audio_bg5.webp"],"firstList":[{"name":"/data/apps/damp/temp/ziptemp/APY230117001_demo1695809214287/APY230117001_demo/056480_demo.jpg","url":"https://bj-oss-datatang-03.oss-cn-beijing.aliyuncs.com/filesInfoUpload/data/apps/damp/temp/ziptemp/APY230117001_demo1695809214287/APY230117001_demo/056480_demo.jpg?Expires=4102329599&OSSAccessKeyId=LTAI8NWs2pDolLNH&Signature=2d2r3%2FYQASEMnhCBsTgJBTu2y74%3D","intro":"","size":0,"progress":100,"type":"jpg"}]}
Vertical Text OCR Dataset with 57,645 Images and Polygon Annotations
ocr dataset
ocr training data
scene text dataset
text recognition dataset
text detection dataset
This dataset contains 57,645 images collected from diverse real-world text scenes, including street scenes, shop signs, billboards, posters, decorative text, art lettering, and magazine covers. The dataset includes Chinese text and a small amount of English text. Each image is annotated with text localization information and transcription labels, including polygon and quadrilateral bounding boxes for vertical and non-vertical text regions. This dataset is suitable for tasks such as scene text OCR tasks and multi-oriented text recognition.
This is a paid datasets for commercial use, research purpose and more. Licensed ready made datasets help jump-start AI projects.
Specifications
Data size
57,645 images, 528,553 bounding boxes
Collecting environment
including street scenes, plaques, billboards, posters, decorations, art lettering, magazine covers etc.
the image data format is .jpg, the annotation file format is .json
Annotation content
vertical -level rectangular bounding box (polygonal bounding box, parallelogram bounding box) annotation and transcription for the texts; non-vertical rectangular bounding box (polygonal bounding box, parallelogram bounding box) annotation and transcription for the texts
Accuracy
The error bound of each vertex of a bounding box is within 3 pixels, which is a qualified annotation, the accuracy of bounding boxes is not less than 97%; The texts transcription accuracy is not less than 97%.
Nexdata provides OCR datasets covering a wide range of data types, including documents, handwriting, invoices, test papers, and forms. The datasets cover diverse languages, layouts, text styles, and real-world scenarios to support OCR model training, text recognition, document analysis, and other document AI applications.
Can Nexdata customize OCR datasets for specific requirements?
Yes. If our off-the-shelf OCR datasets do not fully meet your requirements, Nexdata provides flexible custom data collection, annotation, and quality control services. We can customize datasets based on target languages, document types, handwriting styles, layouts, image conditions, data volume, and annotation specifications for specific OCR applications.
How does Nexdata ensure the quality and scalability of OCR datasets?
Nexdata applies multi-stage quality control throughout data collection, annotation, validation, and delivery. With extensive resources covering documents, handwriting, invoices, test papers, and forms, we can support both large-scale OCR projects and specialized datasets with customized formats, annotation requirements, and quality standards.
Specifications
Sample
Recommended Dataset
