PTZ-Calib: Robust Pan-Tilt-Zoom Camera Calibration

TL;DR

PTZ-Calib proposes a robust two-stage calibration framework for PTZ cameras that combines offline reference-frame calibration via PTZ-IBA with online relocalization to estimate pose, intrinsics, and distortion for arbitrary viewpoints; it additionally supports georeferencing using sparse 2D-3D annotations. The offline stage collects 360-degree reference frames, builds ray-landmark tracks through deep feature matching, and performs incremental bundle adjustment to obtain a consistent local camera model; the online stage relocalizes new views to these references for fast, accurate calibration. Extensive experiments on real (WorldCup) and synthetic datasets demonstrate superior accuracy and robustness over state-of-the-art methods, with efficient online performance and actionable georeferencing capabilities for urban applications. The approach enables practical deployment of PTZ cameras for panorama stitching, localization, and digital twins, with open-source code available for reproducibility.”

Abstract

In this paper, we present PTZ-Calib, a robust two-stage PTZ camera calibration method, that efficiently and accurately estimates camera parameters for arbitrary viewpoints. Our method includes an offline and an online stage. In the offline stage, we first uniformly select a set of reference images that sufficiently overlap to encompass a complete 360° view. We then utilize the novel PTZ-IBA (PTZ Incremental Bundle Adjustment) algorithm to automatically calibrate the cameras within a local coordinate system. Additionally, for practical application, we can further optimize camera parameters and align them with the geographic coordinate system using extra global reference 3D information. In the online stage, we formulate the calibration of any new viewpoints as a relocalization problem. Our approach balances the accuracy and computational efficiency to meet real-world demands. Extensive evaluations demonstrate our robustness and superior performance over state-of-the-art methods on various real and synthetic datasets. Datasets and source code can be accessed online at https://github.com/gjgjh/PTZ-Calib

Figures (6)

• Figure 1: Overview of PTZ-Calib. Given a set of camera images(a), our method can automatically calibrate them in the local coordinate system(b) and further align them geographically using global 3D references(c).
• Figure 2: Pipeline of the two-stage PTZ-Calib method. In the offline stage, the reference images selected from camera images are auto-calibrated with our novel PTZ-IBA algorithm. Once online, our method estimates the camera parameters for arbitrary viewport images accurately and efficiently.
• Figure 3: Comparisons of sports field registration on the WorldCup dataset. The sports field template is projected onto the image in red based on the estimated camera parameters. Misregistrations are pointed out with yellow arrows.
• Figure 4: Undistortion results. Regions to focus on are highlighted.
• Figure 5: Qualitative comparison on the synthetic dataset. The red boxes indicated the regions to focus on.