Vision Based Joint Angle Estimation of Robotic Plasma Medicine Devices
Abstract
This study proposes a vision-based algorithm for joint angle estimation of robotic plasma medicine devices with a steerable tip joint. The algorithm uses ArUco markers of size 4x4 because of their robust and fast detection. An ArUco marker with a square length of 9mm is attached to each of the rigid distal and proximal parts of the robot with respect to the tip joint connecting those parts, and the joint angle is measured using a high-definition (HD) webcam simulating an endoscope camera. The proposed vision-based method can eliminate the need for any additional physical sensors that increase the device size, and the approach is also free of any interference from the electromagnetic field generated by plasma, and the risk of electrical breakdown, which can disturb and potentially damage other sensors with conducting components.
The performance of the ArUco markers is explored in terms of the angle accuracy and precision with respect to other approaches such as electromagnetic (EM) position and orientation sensors and the kinematics model of the joint. Subsequently, the ArUco marker-based vision system is tested with real-time joint angle estimation for the continuous motion of the distal tip. The results from this series of experimental studies serve to validate the effectiveness of the proposed marker-based vision system as a feedback sensing component for the robot-assisted plasma medicine devices.
Citation
Khan, Taimoor Daud (2019). Vision Based Joint Angle Estimation of Robotic Plasma Medicine Devices. Master's thesis, Texas A&M University. Available electronically from https : / /hdl .handle .net /1969 .1 /200729.