Contact us

Features

Introducing functions that streamline operations using AI OCR.

invoiceAgent AI OCR

invoiceAgent AI OCR digitizes documents with a high recognition rate,
minimizing the need for visual checks.

01

5 AI OCR engines

SPA can choose from 5 OCR engines the most suited for the document
to be read. This allows users to use AI OCR while keeping total pay-per-use costs low.

  • More information

    ABBYY is
    a print OCR service by
    ABBYY Japan.

    Unmatched recognition accuracy

    Thousands of companies use ABBYY technologyto process more than 9.3 billion pages *annually

    *Based on ABBYY's internal data

    Preferred OCR

    • ・Used by over 50 million people in more than 200 countries
    • ・Award-winning OCR technology
    • ・An OCR engine that was cultivated by over 20 years of development
  • More information

    Tegaki is an AI OCR service
    by Cogent Labs.

    Original OCR has a 99.2% recognition rate*

    It utilizes a proprietary algorithm that utilizes OCR and deep learning. It can recognize various handwritten characters such as hiragana, katakana, kanji, numbers, alphabets, and symbols.

    *We do not guarantee a 99.22% recognition rate for Tegaki for all handwritten forms for the first reading.

    Highly convenient and able to read a diverse range of characters

    Since both printed and handwritten text can be read by a single service, you do not need to use or select different OCR systems for different forms.

  • More information

    WingArc Data Capture is
    an AI-based OCR engine
    developed by WingArc1st.

    Supports a wide range of Japanese character codes

    In addition to print (alphanumeric and Japanese), handwriting (alphanumeric) is also supported. It supports JIS level 1, JIS level 2, JIS level 3, and JIS level 4 Japanese character codes, as well as supplementary kanji.

    Converts checkboxes and enclosed characters to data

    Convert handwritten checks and enclosed characters. Checkbox recognition rate is highly accurate at 99.5%.

  • More information

    DEEP READ is
    an AI OCR service
    by Edulab.

    Instantly convert handwritingto text using AI

    DEEP READ achieves the highest industry class for reading accuracy by comparing more than 30,000 data from various industries, including finance, education, and government offices.

    On-premise versionavailable

    An on-premise version can be provided if cloud is not available due to security policies.

    Digitize textexcluding unecessary information

    Ruled lines can also be digitized.

  • More information

    CLOVA OCR is
    an AI OCR service by LINE.

    Advanced reading accuracyfrom AI OCR specialized for invoices

    Key information items can be recognized with high accuracy without prior configuration or learning, even for invoices with different formats for different suppliers.

    High throughput and quick response

    Its infrastructure is designed to withstand peak access times due to the processing of huge amounts of invoices. This enables quick response even at the beginning of the month, when processing tends to be concentrated.
    It is stress-free and easy to use.

OCR sample

You can read text according to the characteristics of the items in each document.

Switch AI OCR engine by field

Users can specify the AI OCR engine to be used for each field, even when documents
contain a mix of printed text, handwriting and enclosed characters.

02

OCR specialized for invoices

With the AI OCR specialized for invoices, you do not need to set any reading settings beforehand.
invoiceAgent AI OCR automatically analyzes, recognizes, and digitizes key items from scanned or electronic document invoices.
Not only does this help facilitate the digitization of invoices with different formats sent to business partners and boost operational efficiency,
it also ensures proper storage and compliance with the Electronic Book Storage Act.

*The digitization of detail data is not supported.

03

Data comparison function

Enabling multiple OCR engines allows users to compare OCR results.
If the results are the same, the Verified flag will be set automatically.
Users can fully automate validation and revision tasks with the data comparison function.

*When using pay-per-use OCR, users will be charged for each OCR used.

04

Two types of format definitions

Fixed position type

Set the OCR engine by specifying the areas to be read in advance.
This reduces the risk of false recognition in OCR processing, improves the recognition rate, and reduces the man-hours required for verification.
It only takes a few minutes to register forms (document definition) using mouse operations.
You can also copy, paste, import, and export the registered field settings information.

Fixed item type

You can also use OCR to read text for documents with a free format by specifying the items to be read.

  • To read a single item: Adjacent Text field
  • To read a line: Detail table field
Example of reading an Adjacent Text field
Example of reading a Detail table field

05

Automatic identification and auto-correct function

When documents with different layouts are scanned in a single batch, the contents of the PDF files can be identified and automatically sorted.
Related documents are recognized as a set and then digitized.
Scanned images are corrected for distortion, rotation, and scaling before being they are read,
which prevents reduced OCR recognition rate for documents that were distorted during scanning or camera capture.

06

Other AI OCR features

CSV output of OCR results
Information digitized by OCR can be output as a CSV file. You can also retrieve CSV files and link them to external systems using Web API.
Remarkable development productivity
The Document Definition Configuration tool lets you create a document definition in a few minutes. You can also copy, paste, import, and export field settings information.
False recognition learning function
Improve the accuracy rate by studying the letters that are incorrectly read.

Example:
Before correction: WingArc 1st Inc
After correction: WingArc1st Inc.
OCR statistical data
AI OCR operation log data can be output and visualized as statistical information.
Use the invoiceAgent AI OCR visualization option to measure the effectiveness of OCR.
The recognition rate and work hours for each document can be visualized in an easily understandable manner, which is useful for understanding and improving the current situation.
Skip blank fields
Blank fields that do not contain text can be skipped, reducing OCR costs.
このページのトップへ
Back to Top