Detection vs recognition, the four-step pipeline (detect, align, embed, match), 1:1 verification vs 1:N identification with the false-positive scaling math, the honest field accuracy (cooperative ~99% vs CCTV ~65–85%, never 100%, with demographic differentials), how a face event surfaces via ONVIF Profile M and where it runs, and the legal gate (GDPR Art. 9 · EU AI Act Art. 5 · Illinois BIPA · Texas CUBI) — on one page.
Download free PDF
The handbook · 7 volumes
Fora Soft Video Engineering Handbook
This reference covers the detect, align, embed and match pipeline, 1:1 versus 1:N false-positive math, realistic CCTV face recognition accuracy and the legal gate. Our seven-volume handbook covers the video surveillance stack; start with Volume 6 on detection, tracking and computer vision, then Volume 7 for biometrics regulation and Volume 1 for encoding. Written by Nikolay Sapunov, CEO at Fora Soft.
Evaluating face recognition for CCTV and need realistic accuracy numbers?
Share your cameras and use case, and we will estimate field accuracy and false matches, and check the legal requirements. 250+ video products shipped since 2005.
You get a realistic estimate, technical recommendations and next steps. Free, no obligations.
