Quality Thresholds (Document Image)

Quality thresholds define the minimum capture standards an ID or PoA image must meet before automated verification proceeds. Typical criteria include sharpness (no motion blur), brightness/contrast, glare limits, edge completeness, resolution (e.g., ≥300–400 DPI equivalent), and unobstructed security features/MRZ. Enforcing thresholds reduces false negatives/positives in OCR/ICR, face match, and document forensics, lowering manual review and fraud risk. Modern SDKs perform on-device quality checks with real-time prompts (reposition, remove glare, increase light) before upload. Compliance teams document these thresholds to meet auditability, ensure consistent onboarding outcomes, and protect against acceptance of tampered or unreadable IDs. Thresholds should vary by document type and channel (mobile vs. web) and be tuned using empirical error rates and reviewer feedback.