A prerequisite for using smart camera networks effectively is a precise extrinsic calibration of the camera sensors, either in a fixed coordinate system, or relatively to each other. For cameras with partly overlapping fields of view, the relative pose estimation may be directly performed on or assisted by the video content obtained during scene analysis. In typical conditions however (wide baseline, repetitive patterns, homogeneous appearance of pedestrians), the pose estimation is imprecise and very often is affected by large errors in weakly constrained areas of the field of view. In this work, we propose to rely on progressively stricter constraints on the feature association between the camera views, guided by a pedestrian detector and a re-identification algorithm respectively. The results show that the two strategies are effective in alleviating the ambiguity which is due to the similar appearance of pedestrians in such scenes, and in improving the relative pose estimation.

This paper presents a concept which tackles the pose estimation problem (extrinsic calibration) for distributed, non-overlapping multi-camera networks. The basic idea is to use a visual SLAM technique in order to reconstruct the scene from a video which includes areas visible by each camera of the network. The reconstruction consists of a sparse, but highly accurate point cloud, representing a joint 3D reference coordinate system. Additionally, a set of 3D-registered keyframes (images) are used for high resolution representation of the scene which also include a mapping between a set of 2D pixels to 3D points of the point cloud. The pose estimation of each surveillance camera is performed individually by assigning 2D-2D correspondences between pixels of the surveillance camera and pixels of similar keyframes that map to a 3D point. This allows to implicitly obtain a set of 2D-3D correspondences between pixels in the surveillance camera and their corresponding 3D points in a joint reference coordinate system. Thus the global camera pose can be estimated using robust methods for solving the perspective-n-point problem.

While for static cameras several background subtraction approaches have been developed in the past, for non-static pan/tilt cameras efficient and robust motion detection is still a challenging task. Known approaches use image-to-image registration methods to generate a panorama background model of the scene, which spans a joint pixel coordinate system for later background estimation and subtraction. However, for a real-time panorama-based background subtraction a highly efficient image-to-panorama registration is needed. For this purpose, in this paper a key-frame representation of the panorama image is proposed and a strategy for fast global homography estimation in large panorama images is presented.