Abstract: Reinforced carbon fiber composite skin is generally connected to aircraft skeleton parts through connection holes. The hole position is crucial for ensuring the joint strength. In order to achieve efficient online measurement of hole position on textured composite skin in polluted environment, this paper designs a machine vision device, and image algorithm is also proposed. Accurate extraction of hole edge is realized by rapid positioning of hole area and edge search around hole wall. The hole position is calculated based on image algorithm, laser beam calibration and spot coordinate. The online measurement experiment of hole on skin test piece is carried out. Compared with common image algorithm, the proposed image algorithm has better adaptability. The experiment result shows that measuring error of hole position is less than 0.05 mm, while the system can measure 12 holes per minute.