- TH-OCR SDK
- OCR Text Recognition System
- TH-OCR Minority Document Recognition
Function Introduction
-
1:1 reduction of reconstructed forms
Supports automatic recognition of analyzed forms and 1:1 reduction of reconstructed forms to ensure data integrity
-
OCR Recognize newspapers and magazines with different layouts
OCR Recognizes the contents of newspapers and magazines with different layouts and accurately reproduces the original layout.
-
Support a variety of collection methods
Images acquired through cell phones, tablet PCs, high-frequency cameras, scanners, digital cameras and other means can be recognized.
-
OCR Recognizes multiple languages with wide character sets
Supports recognition of printed simplified Chinese, handwritten simplified Chinese, printed traditional Chinese, handwritten traditional Chinese, rare characters, mixed text of handwritten and printed Chinese, and general English.
-
Supports OCR recognizing and exporting images/files in various formats
Support for recognizing BMP, JPG, JPEG, PNG, PDF and other mainstream image formats; export JSON, TXT, double PDF and other formats of the file
-
Automatically rotate images
TH-OCR text recognition supports 45 °, 90 °, 180 °, 270 ° and other multi-angle picture rotation.
-
Can be integrated with a variety of system applications
Provide standard API interface to realize rapid integration with archive, OA, ERP and other application systems.
-
Multiple deployment methods
Support both B/S end service deployment and PC end SDK integration deployment
-
Support for CPU and GPU hardware configuration deployment
Support CPU and GPU hardware configuration deployment under the same identification performance, saving hardware investment costs.
Product Superiority
-
The Second Prize of National Science and Technology Progress
- TH-OCR Technology won the Second Prize of National Science and Technology Progress, the First Prize of Science and Technology Progress of the Ministry of Education and many other awards
-
Expanding applications
- Support to expand the application with AI scanners and other hardware products to achieve the management and digitization of large amounts of data files
-
Automatically Image Rectification
- Support automatically rectify the tilt image cased by scanning
-
Support layout analysis and layout restoration
- Supports automatic layout analysis of images and 1:1 restoration of layout format and information according to the analysis results.
-
OCR Intelligent Recognition
- Online recognize the text characters location and font size ,can recognize multiple PDF file
-
Support Uncommon Words Recognition
- High Accuracy Recognition a various of uncommon words
-
Remove Underline, Filter Color and Recognize
- Filter image pink and red interference background color
-
Ensure Data Security
- Deploy with enterprise intranet to meet users' requirements for data security
-
Fast OCR Recognition
- Single image can be recognized within a second, recognition accuracy is as high as 99%
-
Wide Range of Application
- State Intellectual Property Office, Geological Survey Bureau, etc., have integrated Sinosecu OCR text recognition into enterprise application system
Application Scenarios
- Electronic archives security management system
- System office software
- Digital processing of archives
- Intelligent Hardware
- Contract Management
- Large model
-
Electronic archives security management system
Wintone TH-OCR character recognition SDK is used in computer terminal security inspection system, electronic document security management system, e-mail prevention and control management system, electronic file security management system and other systems. The ultra-high recognition rate provides technical support for data security and risk prevention.
-
System office software
The OA software office system is an indispensable software tool for enterprises. The Wintone Science & technology TH-OCR text recognition SDK is embedded into the OA office software and installed on the client side together with the OA software to provide end customers with official documents or internal information. Text extraction function, easy to operate, improve overall work efficiency and save labor costs.
-
Digital processing of archives
Connect Wintone Science & technology TH-OCR text recognition SDK with business systems such as information service system, smart office system, and file digitization system, and apply it to the management and digitization of large amounts of data files. By scanning enterprise paper files, it can be exported Double-layer PDF, TXT, WORD, XML, XLS and other format files are convenient for enterprises to edit and find file information, realize digital management of files, and improve efficiency and intelligence.
-
Intelligent Hardware
The Wintone Science & technology character recognition SDK is embedded in the multi-function all-in-one machine, scanner or other intelligent hardware, which can quickly identify and extract characters during document entry and realize the digital management of enterprise information.
-
Contract Management
Contract comparison is based on Wintone Science & Technology TH-OCR character recognition, which automatically extracts the key information in the contract text, identifies the position and size of the word in the original text, outputs the image recognition as an editable file format for storage, realizes computer instead of manual review and comparison, and solves the problem of high time cost and high labor cost of manual review in contract management.
-
Large model
Wintone OCR recognition technology combined with big models can more accurately recognize and classify information by understanding the overall content and structure of documents, thus significantly improving the efficiency of document retrieval and management. For example, the deep learning model can understand complex document layouts, automatically extract key information, handle various text and layout styles, and ensure the quality and editability of the converted electronic text, greatly reducing the time and cost of manual operations.
Case
Online experience
TH-OCR SDK Distinguish
The recommended: size of the image is about 200KB, and the bit depth is more than 24Recommended resolution of scanned image: 300dpi, less than 3m.
- TH-OCR SDK
- Result
- JSON Data