Image Acquiring and Stitching Method for on-site Hydraulic Turbine Blade Robotic Maintenance
In order to improve the performance of on-site robotic repair of hydraulic turbine blades, it is needed to build the blade surface map for detecting, evaluating and localizing the cavitation. In this paper an approach to acquire and stitch images of hydraulic turbine blade surface on site is proposed. A wheeled climbing robot which can climb on three-dimensional complexly shaped surface with good maneuverability is designed to carry out image acquisition. A mosaicking algorithm is developed, including image features extraction, features matching, computing the transform parameters between images and merging them to build a surface map. The system is implemented and experiment is performed. Results validate the feasibility of the developed method.
Hydraulic turbine Map building Climbing robot Image stitching
Zheming Wu Zhenguo Sun Qiang Chen
Department of Mechanical Engineering, Tsinghua University, Beijing, 100084
国际会议
济南
英文
1-8
2012-08-24（万方平台首次上网日期，不代表论文的发表时间）