Calibration of a dynamic camera cluster for multi-camera visual SLAM

Encoderless Gimbal Calibration of Dynamic Multi-Camera Clusters

“…The calibration of the time-varying extrinsics for a DCC was achieved by Das and Waslander [5]; this approach uses the Denavit-Hartenberg (DH) parameters [13] to model the forward kinematics of the actuated link as a serial manipulator, and then estimates the calibration parameters using overlapping viewpoints of a fiducial target. The work in [5] was later extended to perform fully automatic viewpoint selection and calibration using a locally optimal next-best-view approach [6]. Although the existing DCC approaches provide good calibration results, the proposed methods require accurate measurements of the mechanism's joint angles.…”

“…We use a static fiducial marker of known size and assign to it a coordinate frame F t . We first summarize the authors' previous DCC calibration approach [5] which incorporates encoder measurements, and then describe our proposed method which performs the calibration of the DCC without the use of encoder feedback. Finally, we illustrate how a VIO algorithm can be adapted to simultaneously perform online estimation of the unknown mechanism joint angles and the VIO localization state.…”

“…Note that in (4), the kinematic parameter vector, Θ * , consists of the parameters required to define the transformation from the static camera frame to the base frame of the manipulator, the DH parameters of the actuated mechanism, and finally the parameters corresponding to the transformation from the end effector frame to the moving camera frame. The calibration process can also be formulated using a constrained optimization technique, and can be extended to perform a DCC calibration between any combination of static and dynamic camera pairs [5].…”

“…The authors' previous work outlines the calibration of time-varying extrinsic parameters for various DCC configurations, such as one static and one actuated camera as well as two actuated cameras [5]. The DCC calibration was also extended to determine optimal gimbal configurations which locally minimize parameter uncertainty using nextbest-view [6].…”

Geometry-Based Camera Calibration Using Closed-Form Solution of Principal Line

“…C AMERA calibration is a crucial step in many 3D vision applications, such as robotic navigation [1], depth map estimation [2], and 3D reconstruction [3]. Camera calibration establishes the geometric relation between 3D world coordinate system (WCS) and the 2D image plane of camera by finding extrinsic and intrinsic camera parameters.…”