Stability Detection of Building Bearing Structure Based on Bim and Computer Vision

The stability of building load-bearing structure is closely related to the building life. Unstable building load-bearing structure has a great potential safety hazard. The stability detection method of building load-bearing structure based on BIM and computer vision is designed. The computer vision system for image acquisition of building load-bearing structural components is designed, which is composed of CCD camera, image acquisition card, light source and UAV. The secondary development technology of Revit is applied to establish the overall structural information model, and the influencing factors of structural stability are introduced into the information model. Build a stability detection network model based on YOLOv4 network architecture, and implement the stability detection of building load-bearing structures. The test results show that the accuracy rate of the design method is higher than 0.920, the recall rate is higher than 97%, and the test time for new and old buildings is 106.85 min and 75.41 min respectively under different confidence levels.