A new detector named Alignment Detector is added to the system. This change applies to customers using the Detect API or Detect product through Ocrolus Dashboard. Currently, this detector only applies to bank statements, pay stubs, and W-2 PDFs.
The Alignment Detector analyzes the PDF and finds the expected alignment of fields within the document based on their orientation and positioning. The expected alignment is highlighted in gray. The fields that do not align with the expected alignments are highlighted in a red box. The specific text value that is misaligned is highlighted in dark red.
Fraud Analysts can use this feature to check for misalignments and font edits that are hard to spot with the naked eye. In addition to the existing detectors and visualizations, this adds to the Detect capabilities to catch and present fraudulent documents for reviewing and preventing bad loans from being allowed.
To learn more about the new alignment detector, see the Understanding signals tab on the Understanding signals page.