The construction schedule of water conservancy projects is an important matter of great concern to the construction parties.Based on BIM technology and image recognition technology,this paper proposed a virtual and real combining method for pouring height recognition of sluice piers,so as to realize the intelligent construction of sluice projects.Through BIM technology modeling,virtual image data sets containing different pouring heights of gate piers were generated.The object detection algorithm YOLOv5 of deep learning was used to train the virtual data set,and an intelligent model of accurately identifying piers with differ-ent pouring heights was obtained.The real pier models with different pouring heights were built in the laboratory,and the real ima-ges of the piers with different heights were captured by the camera and verified by the established intelligent recognition model.The results showed that the model can accurately identified the pouring heights of piers in actual scenes.This method can be wide-ly used to identify the actual construction progress of buildings in hydraulic engineering,and provide effective model support for the construction of digital twin hydraulic engineering.
维普
万方数据