Text-to-pass forms/Notes/document OCR recognition engine

Requirements background:

In the economic activities produced a large number of paper forms of document documents, in order to achieve the huge number of large, wide variety of paper archives electronic storage, the need for document scanning, data entry, manual proofreading and other work. Traditional manual Input method, users need to invest a lot of manpower cost and time cost, not only raised the operating costs, but also difficult to improve the speed of entry, the error rate is difficult to reduce, to improve business processing time, improve the quality of enterprise service brings a lot of negative impact.
Compared to the traditional manual input method, OCR (Optical Character recognition, optical character recognition) Intelligent input has a strong advantage, OCR recognition speed far faster than manual entry, saving a lot of human resources, optimize the allocation of resources, Assign people to more meaningful work. OCR technology can provide users with high-efficiency, low-cost data acquisition solutions, so as to provide strong support for rapid business development.
Document/document recognition engine using Tsinghua TH-OCR core technology, built-in rich and efficient image processing technology, providing a flexible and rich API interface, is an excellent data acquisition solution, is the best choice for users to improve business efficiency and reduce operating costs.

Performance parameters:

Identification Type

Printed: Chinese (Simplified and Traditional), English, digital
Handwriting: Chinese (Simplified and Traditional), English, digital
Specific format: Date type, amount type, flow details, etc.
Symbol Class: Identification code, barcode (one-dimensional, two-dimensional), CheckBox, ╳, ▇, etc. customizable symbols
Other: Image slicing, red chapter detection, etc.

Recognition speed

The recognition speed of single image is less than 1 seconds;

Recognition rate

99.96% accuracy for the identification of printed characters with neat layout
For the standard handwritten numerals, English recognition, the correct rate of up to 98.5%
For the standard handwritten Chinese character recognition, the correct rate is about 85%;

Image processing

Automatic direction reversal
Tilt correction

The following OCR products may meet your needs (product link):
    1, financial Invoice data recognition engine (SDK)
    2, table Form automatically identifies the development package
    3, mobile platform OCR recognition engine
    4, " Express Pass "ticket scanning recognition system
    5, image OCR recognition platform
    6, Wen Tong Bank Note Identification Development package
    7, Check, acceptance recognition engine (SDK)
    8, document Image Intelligent Classifier

Text-to-pass form/ticket/document OCR recognition engine