A New Machine Vision Method for Target Detection and Localization of Malleable Iron Pipes: An Experimental Case

Malleable iron pipes are widely used in construction, manufacturing, aerospace, and many other fields. Cast malleable need to be treated flat meet the needs of different shapes sizes. This process is usually completed manually, which low efficiency subject potential safety risks. To solve this problem, a machine vision method proposed detect localize pipes. Point cloud images obtained by Random Sample Consensus (RANSAC) algorithm, precise matching Iterative Closest (ICP) algorithm obtain more accurate positions, so as realize robot grasping. The grasping experiments with same specifications were carried out using specially designed experimental platform. results show that can identified effectively corresponding success rate than 85 %. target detection localization three-dimensional (3D) position improve efficiency, provided certain theoretical basis guiding significance production practice.