A Novel 3D Seam Extraction Method Based on Multi-Functional Sensor for V-Type Weld Seam

Abstract

To meet the current flexible modern manufacturing mode, based on the idea from coarse to fine, a novel three-dimensional(3D) seam extraction algorithm is proposed in this paper which includes the global rough extraction and the local fine positioning. Firstly, a multi-functional vision system based on Digital Light Processing(DLP) projector is designed to adapt different 3D measurement tasks. Secondly, a global sensor based on DLP projector is constructed which is used to acquire the whole weld seam. Meanwhile, a new feature extraction algorithm based on 3D information of work pieces is proposed. Thirdly, a line structured light sensor based on DLP projector is to serve as local sensor. Based on the results of rough extraction, the close-range laser scanning is to improve the measurement precision. Finally, the 3D path model and pose estimation of weld seam are done which could be applied into the 3D path teaching of welding robots. The experiments show that the proposed algorithm could well finish the high-precision 3D seam path planning through the "Global-Local" method.