Doctr OCRInstall

Summary

Extract text from images by performing OCR on binary image data from a specified property, returning plain text, structured OCR data, or both.

Inputs

• Binary Property (required) — Name of the binary property containing the image to process.
• Output Format (required) — Select what data to return from OCR: plain text only, structured OCR data only, or both.
• Options - Python Executable Path — Optional path to the Python executable running the OCR process (default is 'python3').

Output shape

a list of JSON objects each matching one input item, containing extracted OCR text and/or structured data according to the selected output format

Outputs are paired with input items. On failure, errors can be included in the output JSON if "continue on fail" is enabled; otherwise, the node stops execution with an error. The node requires the input item to contain binary data with the specified property name and depends on an installed Python 3 environment with doctr.

Examples

Example 1: Input items contain binary image data in the 'data' property. The operation is configured to output both plain text and structured OCR data. The output items hold JSON with keys 'plainText' and 'structuredData' representing the extracted content.

Binary Property: data; Output Format: both; Options: default python3 executable.