TesseractInstall

Summary

Recognizes text in an image or PDF from an incoming binary field, optionally limiting OCR to a rectangle. Returns extracted text and confidence for each processed image, subject to the minimum-confidence filter.

Inputs

• Input Image Field Name — Binary field containing an image or PDF; defaults to data.
• Detect on Entire Image? — Defaults to true. If false, set Top Y, Left X, Width, and Height to define the OCR rectangle.
• Options → Language — Tesseract language code; defaults to eng.
• Options → Page Segmentation Mode — Controls how Tesseract segments the page; defaults to Single Block.
• Options → Resolution — Optionally force a DPI value; defaults to automatic resolution detection.
• Options → Character Lists — Optionally configure allowed or disallowed characters.
• Options → Timeout — Maximum processing time per image in milliseconds. A timeout raises an operation error; with Continue On Fail enabled, the item is returned with the error.
• Options → Resize Factor — Scales the input image before OCR; defaults to 100, which keeps it unchanged.
• Options → Include Binary Output — Defaults to true. When enabled, includes the processed image under binary field ocr; when disabled, clears binary output.
• Options → Minimum Confidence — Discards results below this confidence threshold; defaults to 0.

Output shape

One output item per processed image, each with JSON fields text and confidence.

Image inputs produce one result per input image; PDFs produce results for each supported embedded image. Each result retains the input binary fields and adds the processed image as ocr when binary output is enabled. Unsupported binary MIME types raise an error. With Continue On Fail enabled, failures are returned as error items.

Links

• n8n community nodes installation guide