Visual Expansion and Real-time Calibration for Pan-tilt-zoom Cameras
Assisted by Panoramic Models
Abstract
Pan-tilt-zoom (PTZ) cameras, which dynamically adjust their field of
view (FOV), are pervasive in large-scale scenes, such as train stations,
squares, and airports. In real scenarios, PTZ cameras are required to
quickly make decisions informed about where to direct its focus through
contextual cues from the surrounding environment. To achieve this goal,
some researches project camera videos into three-dimensional (3D) models
or panoramas and allow operators to perceive spatial relationships.
However, these works face several challenges in terms of real-time
processing, localization accuracy, and realistic reference. To address
this problem, we propose a visual expansion and real-time calibration
for PTZ cameras assisted by panoramic models. We attempt to meet the
demand for real-time processing with a motion estimation model for a PTZ
camera, to improve calibration performance of PTZ images with only two
feature point pairs, and to provide a realistic environmental context
through a panoramic model. We verify our methods on both public and our
self-built test scene. It can be seen from the experimental results that
our method can exhibit impressive accuracy and efficiency.