Layered OCR

Performs OCR using multiple OCR Layers, merging the results into a single output document.

Remarks

Overview

The Layered OCR engine enables advanced recognition workflows by combining the results of multiple OCR operations, each defined by its own OCR Profile and optional extraction logic. This approach is ideal for documents with mixed print types, complex layouts, or where targeted repair of OCR output is required.

How It Works

• The 'Main OCR Profile' property specifies the primary OCR Profile used to generate the base OCR output.
• The 'Layers' property allows you to define additional OCR Layers, each with its own OCR Profile and optional extractor for targeted repair or enhancement.
• Each layer can supplement or repair the main output by focusing on specific regions, print types, or error patterns.
• Results from all layers are merged, with repair logic applied as needed.

Usage Scenarios

• Documents containing both standard text and specialized print types (such as MICR, handprint, or OCRA).
• Structured forms where field labels, table headers, or other background elements require enhanced recognition.
• Automated repair of common OCR errors using extractors to identify and correct problematic regions or values.

Configuration & Best Practices

• Configure the 'Main OCR Profile' for general recognition; add layers only for targeted repair or enhancement.
• Each layer's extractor must meet special requirements for OCR repair. If repair is not working as expected, check the log for error messages and review extractor configuration.
• Keep the number of layers minimal to optimize performance and maintainability.
• Test your configuration on representative samples and review merged output for accuracy.

Properties

Name	Type		Description
Main OCR Profile	OCR Profile	►	The OCR Profile used to establish the base OCR output for the document. EditorNode Reference Editor Requiredtrue The 'Main OCR Profile' defines the primary recognition strategy. It is executed first, and its results form the foundation for all subsequent repair or enhancement performed by additional layers. Configuration Guidance: Choose an OCR Profile that best matches the general content and print type of your documents. Use additional layers only to address specific recognition challenges not handled by the main profile. For more information on configuring OCR profiles, see the documentation for OCR Profile.
Layers	OCR Layer[]	►	Defines one or more OCR Layers which supplement or repair output from the 'Main OCR Profile'. EditorObject Collection Editor ConverterCollection Converter The 'Layers' property allows you to specify additional OCR Layers, each with its own OCR Profile and optional extractor. Layers are processed in order, and their results are merged into the main output. Typical Uses: Targeting specialized print types (e.g., MICR, handprint) with dedicated OCR Profiles. Repairing or enhancing recognition of specific fields, labels, or regions using extractors. Configuration Notes: Each layer must specify an OCR Profile. If the profile matches the main profile, only extraction/repair is applied. The extractor on each layer must meet special requirements for OCR repair. If repair fails, review the log for error messages. The order of layers can affect the final output; arrange layers to prioritize the most critical repairs or enhancements. For more details, see the documentation for OCR Layer.

See Also

OCR ProfileOCR Layer

Used By

OCR Profile