Template

Available in Classic and VPC

In CLOVA OCR > Domain of the NAVER Cloud Platform console, you can create and manage Template domains that support template builder, which can directly specify analysis areas, extract recognized values, test them, and then send results. This guide describes how to create and use Template-type domains. For more information about how to view, delete, and copy a domain list, and how to integrate with an API Gateway, see Domain.

Create domain

The following describes how to create a domain whose service type is Template.

1. From the Region menu on the NAVER Cloud Platform console, click and select the region you're using.
2. From the Platform menu, click to select between VPC and Classic.
3. Click Services > AI Services > CLOVA OCR in order.
4. Click the Domain menu.
5. Click the [Create domain] button, and then click the [General/Template] button.
6. When the Create domain pop-up window appears, enter the required information.
   • Domain name: enter a name between 1 to 50 characters using Korean letters, English uppercase and lowercase letters, Chinese letters, Japanese letters, numbers, and full/half-width special characters (& 。_ - × 々), and click the [Check] button to check for duplicates.
   • Domain code: enter a code between 1 to 50 characters excluding spaces using English uppercase and lowercase letters and special characters (.?&。_-×々), and then click the [Check] button to check for duplicates.
   • Supported language: click to select the type of language supported by the domain.
   • Service type: click to select Template.
   • Recognition model: click to select the recognition model applicable for the domain.
   • Service plan: click to select the service plan applicable for the domain.
7. Click the [Create domain] button.

Run template builder

Once you create a Template domain, you can run the template builder. Template builder is a web console that allows you to create and test templates for extracting text from images and documents. It provides various features required by Template OCR. Template builder consists of the following feature menus.

• Create template: create main template, manage main sample list by template subject to analysis
• Test and analyze: test and check results for templates/models whose analysis area is set, view OCR statistical data by service plan
• Component: register specialized words and synonyms in the glossary and apply them to templates, manage glossaries
• Settings: view domain information, set external connection, set verification integration, set permission management
• Manage deployment: deploy beta template, deploy service

Create template

In Create template, you can create the main template, and manage the main sample list by template subject to analysis.

Template list

The following describes how to create a main template and add samples to manage it.

1. From the Region menu on the NAVER Cloud Platform console, click and select the region you're using.

2. From the Platform menu, click to select between VPC and Classic.

3. Click Services > AI Services > CLOVA OCR in order.

4. Click the Domain menu.

5. Click the [Template builder] button of a Template domain.

6. Click the Template list menu of Create template.

7. Click the [Create template] button.

8. Enter the template name in Template basic information, and then click the [OK] button.

   • Template name: enter a name between 1 to 30 characters using Korean letters, English uppercase and lowercase letters, Chinese letters, Japanese letters, numbers, and full/half-width special characters (?, &, 。, _, -, ×, 々).
   • Recognition model: check the type of recognition model applied to the selected Template Domain.

9. Click the [OK] button.

10. In Set representative sample, upload sample images for the improvement of cognition performance.

    • Image file upload conditions

      • File size: up to 20 MB (for PDF and TIFF files, up to 50 MB)
      • File format: jpg, png, pdf, tiff (supports up to 5-page PDF and TIFF files)
      • File resolution: 150 dpi or higher based on A4, a minimum resolution of 10 px up to 1960 px on long axis recommended
      Note

      High-quality images of 150 dpi or higher, with little distortion and damage, are recommended for images entered through the API.

11. When the uploaded image sample appears, specify the title area of the document.
    

    • Specify title area: upload the main sample image and drag the rectangle for specifying the area
      • Clicking an already specified area enables the rectangle for specifying the area, and clicking cancels the area specification
      • Clicking an already specified area enables the rectangle for specifying the area, dragging the dotted line with the cursor resets the area, and clicking saves the reset area

12. Enter a sample name to be used for the template classification of the document in Main sample name, between 1 to 30 characters using Korean letters, English uppercase and lowercase letters, Chinese letters, Japanese letters, numbers, and full/half-width special characters (?, &, 。, _, -, ×, 々), and then click the [OK] button.
    

    • Clicking the [OK] button saves it and displays the notification window for registering synonyms.
    • Even after clicking the [OK] button, the main sample name can be modified by clicking the [Edit] button.
    • [Manage synonyms] button: by clicking this button, you can register words that are recognized similarly to the main sample name to improve the classification premium. When entering multiple synonyms, separate them with commas.
      
      • <example> Local tax bill, local tax, local tax invoice
      • The main sample name of Manage synonyms can't be changed. You can only register and edit synonyms.

13. Set individual fields you want to analyze by specifying them in Specify analysis field.

    • When you click Add field
      • Specify analysis area: when the rectangle for specifying a new field appears, adjust the size of the rectangle to set it.
        
        • Clicking an already specified area enables the rectangle for specifying the area, and clicking cancels the area specification
        • Clicking an already specified area enables the rectangle for specifying the area, dragging the dotted line with the cursor resets the area, and clicking saves the reset area
        • Clicking copies the field name (original name_#) and settings, and a new field specification box appears below the existing field.
        • You can delete the field by clicking the [Delete] button of the field box.
    • When you click Multi-box (Premium only)
      
      • If input boxes are separated, as in bank account numbers or mobile phone numbers, you can specify multi-boxes to recognize and extract characters inside individual boxes.
      • If you set the field area outside the frame of a multi-box, it only recognizes the characters inside the box you wish to extract. (It does not recognize the multi-box frame.)
      • Specify analysis area: when the rectangle for specifying a new field appears, adjust the size of the rectangle to set it.
        • Clicking an already specified area enables the rectangle for specifying the area, and clicking cancels the area specification
        • Clicking an already specified area enables the rectangle for specifying the area, dragging the dotted line with the cursor resets the area, and clicking saves the reset area
        • Clicking copies the field name (original name_#) and settings, and a new field specification box appears below the existing field.
        • You can delete the field by clicking the [Delete] button of the field box.
    • When you click Check-box (Premium only)
      
      • True is returned if unstructured characters like V, X, O, and · are recognized in checkboxes such as selection boxes or radio buttons, and False is returned if not recognized.
      • Check marks can be recognized through OCR in various documents such as surveys and product subscription documents.
      • When specifying a checkbox, you need to set the field so that the frame of the input box is visible, enable recognition of only the checkbox without other characters or shapes in the specified area, and align the frame of the checkbox and field area box as much as possible.
      • It may not be recognized properly in the following cases: if unstructured characters are located outside the checkbox area, if they are blurry or erased, if they are duplicated in the same box due to edition, or if unstructured characters and guide characters have already been entered in the checkbox for the customer to check off.
      • Specify analysis area: when the rectangle for specifying a new field appears, adjust the size of the rectangle to set it.
        • Clicking an already specified area enables the rectangle for specifying the area, and clicking cancels the area specification
        • Clicking an already specified area enables the rectangle for specifying the area, dragging the dotted line with the cursor resets the area, and clicking saves the reset area
        • Clicking copies the field name (original name_#) and settings, and a new field specification box appears below the existing field.
        • You can delete the field by clicking the [Delete] button of the field box.
    • Unrecognized: if needed, in the field area set as the analysis range, you can specify a certain area other than text input and handwriting areas as unrecognized.
      • Method of specifying the area to be processed as unrecognized in an existing field by clicking Add field
        • Text in the area set as unrecognized is not extracted.
      • Method of using the Masking feature supported when setting general fields (Value type All).
        • It's not supported for multi-boxes, checkboxes, and Value type Numeric.
        • If the instruction text in the field is written as background text, the unrecognition of the guide text (background text) or watermark is not supported.

14. Enter the name for the added field and set related options.

    • Enter field name: enter in Field name using Korean letters, English uppercase and lowercase letters, Chinese letters, Japanese letters, numbers, and full/half-width special characters (?, &, 。, _, -, ×, 々).
      • The field name can be edited freely in the respective field.
    • Set the field: click
      • Set value type and terms: when field or multi-box is clicked as an individual field
        • Value type: when you select All, a mix of text and numbers can be recognized, and when you select Numeric, only numbers can be recognized. By specifying the format of values to be entered in the field, you can extract only specific values.
        • Register terms: you can register a glossary containing domain-specific words and synonyms. When a glossary is registered, only the representative terms included in the glossary are recognized. Synonyms, upon identification, are substituted by representative terms to be displayed. It is processed as a failure when words other than representative terms and synonyms are recognized.
      • Text conversion settings for checkbox recognition result: when checkbox is selected as an individual field
        • Returned values of True (Y) or False (N) for the checkbox are substituted and displayed with specific characters.
          

15. If necessary, set the alignment of added fields, or set the combination of field analysis results.

    • Align field: select fields to align by dragging the cursor while holding down the Shift key, click Align left or Align top of Align field.
      
    • Combine analysis result: click the [Combine analysis results] button to combine OCR extraction values of individual fields to output a new value.
      • Processed OCR result values can be extracted according to the fields to be combined and the output order.
      • When entering a fixed text, "output value for each field + fixed text" are displayed together.

16. Click the [Save] button.

17. Proceed with necessary management tasks in the Template list.
    

    • Change main sample image: click the [Edit] button of Edit main sample from the main sample list.
      • When changing the main sample image, all of the previously saved contents are deleted.
    • Delete template: click the template you want to delete from the main sample list, and then click the [Delete] button.
    • Verify template: to proceed with verification of each template by linking your own verification server, click the [Register] button of the Verification feature in the main sample list.

Sample list

You can check and download the main image sample information registered for each template. The following describes how to view and download the sample list.

1. From the Region menu on the NAVER Cloud Platform console, click and select the region you're using.
2. From the Platform menu, click to select between VPC and Classic.
3. Click Services > AI Services > CLOVA OCR in order.
4. Click the Domain menu.
5. Click the [Template builder] button of a Template domain.
6. Click the Template list menu of Create template.
7. When the template list appears, check the information, and download the image sample if necessary.
   • View uploaded main sample image: click the individual sample number.
   • Download uploaded main sample image: click from Download of the template list.

Test and analyze

In Test and analyze, you can upload image files to extract text from, and then test or download the analysis result. You also can check the statistical data for the number of daily API calls for a service plan, external verification results by template, and recognition results.

Test

The following describes how to upload image files to extract text from, and then test or download the analysis result.

1. From the Region menu on the NAVER Cloud Platform console, click and select the region you're using.
2. From the Platform menu, click to select between VPC and Classic.
3. Click Services > AI Services > CLOVA OCR in order.
4. Click the Domain menu.
5. Click the [Template builder] button of a Template domain.
6. Click the Test menu of Test and analyze.
7. From Test condition, click to select a test condition between Beta and Template analysis.
   • Beta: you can test automatic template classification and analysis area in the same environment as service deployment. Beta testing and service deployment becomes available after beta deployment.
   • Template analysis: you can test templates and models with set analysis areas.
8. Upload the data to be tested to Test data.
9. Check the result from Test result.
   
   • View test result: the analysis result is provided in a table.
   • Download test result: click to select between csv and json, and then click the [Download] button.

Indicator

The following describes how to check the statistical data for the number of daily API calls for a service plan, external verification results by template, and recognition results.

1. From the Region menu on the NAVER Cloud Platform console, click and select the region you're using.
2. From the Platform menu, click to select between VPC and Classic.
3. Click Services > AI Services > CLOVA OCR in order.
4. Click the Domain menu.
5. Click the [Template builder] button of a Template domain.
6. Click the Indicators menu of Test and analyze.
7. Set the search period and the template.
   • Search period: up to 90 days can be set.
8. Check the search results.
   • Number of API calls: the number of API calls during the set period for the service plan in use.
   • Verification result: it refers to the external verification result by template and the number of verifications for the domain. When you hover the mouse over the verification result graph, the numbers of VALID, UNCHECKED, ERROR, and INVALID are displayed.
   • Number of template recognition failures: if there is no recognition result because no template was matched, it is counted as a template recognition failure. The number of failures during the set period is displayed.
9. To download the search result as an xls file, click the [Download integrated indicators] button.

Manage glossaries

In Manage glossaries, you can register specialized words and synonyms for the domain in the glossary and apply them to templates. If you specify specific fields to only recognize words from the glossary, or if registered synonyms are recognized, then the result is substituted with representative terms and displayed. The following describes how to set up a glossary.

1. From the Region menu on the NAVER Cloud Platform console, click and select the region you're using.
2. From the Platform menu, click to select between VPC and Classic.
3. Click Services > AI Services > CLOVA OCR in order.
4. Click the Domain menu.
5. Click the [Template builder] button of a Template domain.
6. Click the Glossary menu of Manage glossaries.
7. Click the [Create glossary] button.
8. When the Create glossary pop-up window appears, enter the required information.
   
   • Glossary name: enter a unique glossary name that isn't a duplicate.
   • Representative terms: enter the representative terms that make up the glossary.
   • Synonyms: enter the synonyms of entered representative terms. When entering multiple synonyms, distinguish them by using a comma (,). <example> local tax bill, local tax, local tax invoice
9. Click the [OK] button.

Settings

In Settings, you can check the domain information, or set options such as external linkage and permission management. You can check and configure them as follows:

1. From the Region menu on the NAVER Cloud Platform console, click and select the region you're using.
2. From the Platform menu, click to select between VPC and Classic.
3. Click Services > AI Services > CLOVA OCR in order.
4. Click the Domain menu.
5. Click the [Template builder] button of a Template domain.
6. Click the [Set] button.
   
7. Click the desired tab menu to check or set the required information.
   • When clicking the [Domain information] tab menu: view the domain and template information
   • When clicking the [API Gateway integration] tab menu: set the API Gateway integration
     • For detailed instructions, see Domain.
   • When clicking the [External verification server] tab menu: click the [Create] button to manually set and manage the address to which OCR results will be sent.
     
     • Enter the name in the ${result transfer name} format.
     • Enter a URL that begins with https:// in the URL field.
   • When clicking the [Manage permissions] tab menu: add new admins and grant permissions
     • When you click the [Create URL] button, an Invite URL is created, with which you can grant permissions.
     • The Invite URL is deleted if you click the [Delete] button.

Manage deployment

Beta deployment is the final storage feature reflected in the model. You can learn the samples, analysis areas, and glossaries that have been applied to templates. Beta testing and service deployment can be performed after beta deployment. The following describes how to manage deployment.

1. From the Region menu on the NAVER Cloud Platform console, click and select the region you're using.
2. From the Platform menu, click to select between VPC and Classic.
3. Click Services > AI Services > CLOVA OCR in order.
4. Click the Domain menu.
5. Click the [Template builder] button of a Template domain.
6. Click the [Manage deployment] button.
   
7. When the template list appears, check the latest status of each template from Current deployment status.
8. Click and select the template to deploy, and then click the [Deploy beta] button.
   • Beta testing becomes available in the Test menu after the completion of beta deployment.
   • Information edited after testing the analysis result in the Test menu is applied through a new beta deployment.
9. Click the [Deploy service] button to perform a service deployment of the latest beta deployment template that completed the beta test.

Integrate with API Gateway

The following shows how to integrate the API Gateway to set the Invoke URL for Template OCR.

1. From the Region menu on the NAVER Cloud Platform console, click and select the region you're using.
2. From the Platform menu, click to select between VPC and Classic.
3. Click Services > AI Services > CLOVA OCR in order.
4. Click the Domain menu.
5. Click the [Template builder] button of a Template domain.
6. Click the [API Gateway integration] tab menu, and then click the [Integrate] button.
7. When the API Gateway integration pop-up window appears, set up the required information.
   • For detailed integration methods, see Integrating with API Gateway of Domain.

Call API

Once you complete the Invoke URL settings of Template OCR, you can enter images by calling the API provided by CLOVA OCR, and get the recognized value. For more information about how to call APIs, see the following guide:

• CLOVA OCR API guide
• Template OCR API call examples