Multi-camera tracking improves by calibrating homographies on the fly
On-the-Fly Homographies Calibration for Multi-Camera Tracking
Computer Vision and Pattern RecognitionArtificial Intelligence
Summary
Tracking people or objects across multiple cameras usually needs careful setup that takes a lot of work and space, which isn’t always possible, especially in big areas or where privacy rules forbid recording videos. The authors created a system that adjusts camera positions automatically and continuously by using only simple position data from live detections, without any videos or heavy computing. This means the system can keep tracking accurately even if cameras move or the scene changes, all while protecting privacy and working efficiently.
What this means in practice
- •For security operations teams: Maintain accurate, privacy-compliant tracking across large camera networks without manual recalibration or video recording.
- •For smart city engineers: Deploy scalable multi-camera tracking infrastructure that adjusts automatically to environmental changes and camera movements in urban areas.
Authors
David Voihanski, Mor Sinai, Ben Zion Bobrovsky
Abstract
Precise multi-camera tracking traditionally relies on rigorous 3D site calibration, yet this requirement is often operationally impossible in large-scale deployments. Privacy regulations frequently prohibit recording video for offline calibration; limited bandwidth precludes synchronizing high-resolution streams from hundreds of cameras; and covering immense physical sites with calibration targets is logistically infeasible. We present a multi-camera homography calibration system designed to overcome these barriers through "on-the-fly" geometric refinement. Starting from coarse manual homographies, we introduce a centroid-based projection optimization (PO) that continuously aligns the ground-plane geometry using live detection streams. Because PO operates asynchronously on already-transmitted, lightweight metadata, it adds zero computational latency to the real-time tracker. This allows the system to adapt automatically to camera movements or environmental changes without human intervention. This optimized geometry feeds a multi-camera bird's-eye-view (BEV) tracker that fuses detections and unifies trajectories across zones. Crucially, by operating strictly on live anonymous metadata, our solution ensures a privacy-safe, zero-overhead, and resilient tracking pipeline that maintains global consistency in dynamic environments where static, recorded-video calibration is impossible.