Motion Estimation and Control of a Flexible Serpentine Robot by Motion Stereo and Distributed Sensors

研究了主动瞄准镜摄像机内部传感器对运动和电缆形状的估计。仅使用相机图像和运动传感器的方法可以在室内环境中估计6自由度运动，但碎石环境中由于光线的运动使估计困难。通过将无气味卡尔曼滤波应用于柔性电缆的ANCF动态模型，提高了对稀疏传感器和大干扰的估计精度。

Estimation of motion and cable shape was studied using internal sensors of Active Scope Camera. By methods using only camera image and motion sensors, 6-DOF motion could be estimated in room environment, but rubble environments made the estimation difficult because of motion of the lights. By applying an unscented Kalman filter to a dynamic model of flexible cables by ANCF, the accuracy of estimation was improved for sparse sensors and/or large disturbance.