X-Ray-Based Three-Dimensional Shape Reconstruction Method for a Continuum Manipulator

The utilization of visual sensors for three-dimensional (3D) reconstruction of continuum manipulators is a well-established technique, offering essential support for precise control and accurate state estimation. This approach has been widely adopted across various fields, including surgical robotics, bionic systems, and industrial automation. However, the effectiveness of visual sensors is substantially constrained in environments characterized by occlusions or spatial limitations, where accurate reconstruction becomes challenging. To overcome these limitations, we propose an advanced 3D reconstruction method leveraging X-ray imaging, enabling robust shape reconstruction and automated tracking of continuum manipulators, even under occluded conditions. In the X-ray-based 3D shape reconstruction method, the intrinsic and extrinsic parameters of the computed tomography (CT) device are calibrated. Besides, the X-ray images are processed with identification and segmentation, centerline detection and Bézier curve fitting. Finally, the 3D shapes of a continuum manipulator are accomplished by matching the two images acquired from two distinct CT device rotation angles and reconstruction. The measurement experiment has demonstrated the accuracy of 3D shape reconstruction achieved 0.5383 mm. The shape reconstruction experimental results also showed the effectiveness of the X-ray-based 3D shape reconstruction method.