Abstract: Through automatic extraction of on-board point cloud data, a large number of three-dimensional characteristic data can be quickly obtained. Using synchronous image data can not only improve the accuracy and efficiency of target interpretation, but also obtain the attribute information of the target object, which plays a very important role in production activities such as quality inspection, component attribute investigation and 3D modeling. In this paper, an automatic image segmentation technology corresponding to the feature data of vehicle-mounted point cloud is proposed. The relationship between the image and the feature information is judged by the position and attitude of the image, so as to realize the optimal image automatic selection and segmentation, generate the appropriate local orthographic image, and mark the correct position of the feature information in the new image.

Key words: vehicle-mounted mobile measurement system, point cloud, 3D characteristic data, image data, automatic image segmentation