Is My PDF Redaction Safe?

PDF pages are assembled from text, images, paths, and other drawing operations. A later dark rectangle can make earlier text disappear on screen while leaving that text extractable. The scanner compares supported text geometry with later opaque overlays and reports a possible incomplete redaction when the structures overlap.

A critical finding identifies a supported high-confidence pattern; it is not a legal conclusion and should be checked against the page and the redaction workflow. Normal page design can contain dark areas, so the detector also uses paint order, overlap, opacity, color, annotation position, and image structure to reduce common false positives.

White-on-white covering is a real concealment technique, but ordinary forms and document templates also draw white rectangles. The scanner suppresses known widget appearances, repeated multi-page layout elements, and visible duplicate text patterns. A remaining unique white overlay is still reported as a warning rather than proof of failed redaction.

Review the source of the document and how the cover was created. If someone used a shape, drawing tool, or background color instead of a dedicated redaction command, recreate the output with a method that removes the underlying information and then inspect the result again.

An image-only page has no extractable text geometry for this check to compare with an overlay. The report therefore marks redaction checking as unavailable or partial when some pages lack a text layer. It does not infer safety from the absence of a finding on those pages.

The scanner may miss image-based redactions, OCR-only content, clipping masks, arbitrary complex overlays, unusual transparency behavior, and unsupported PDF structures. It does not provide compliance certification or establish that a document is legally sufficient for release.