Intelligent detection method for reinforcement spacing of prefabricated components based on RGBD image

Abstract

The application discloses a prefabricated component steel bar spacing intelligent detection method based on an RGBD image, which comprises the steps of data acquisition, steel bar pixel semantic segmentation, steel bar point cloud extraction, steel bar point cloud data enhancement and steel bar spacing automatic calculation, etc., wherein first, a depth camera is used to collect steel bar RGBD images of a prefabricated component, then a trained InternImage semantic segmentation network is used to segment steel bar pixels in the steel bar RGBD images to obtain RGB image data containing only steel bar pixels; then, steel bar point cloud data is extracted based on the internal parameters of the depth camera; then, a DBSCAN clustering algorithm is used for processing to obtain point cloud data of each steel bar; and linear point cloud screening, low-dispersion point cloud screening and point cloud denoising are gradually performed through steel bar point cloud data enhancement; finally, a KD-Tree is used to establish point cloud efficient indexing, and the minimum spacing between each steel bar point cloud is calculated based on a point cloud processing algorithm to obtain the spacing of each steel bar. The effect is that the prefabricated component steel bar spacing intelligent detection can be conveniently, quickly and accurately realized.

Images