Stereovision monitoring for entire collapse of a three-story frame model under earthquake excitation

During the stage of structural collapse, conventional contact sensors are usually taken away from structure to protect sensors, and the entire displacement curves of structural collapse cannot be acquired. To deal with this problem, a stereovision-based displacement measurement method is used to monitor the entire structural collapse. To improve measurement accuracy, a three-step calibration algorithm based on solid circle target is proposed to calibrate stereovision system in this paper. The Digital Image Correlation method is employed to perform sequential image matching of circular target on structural surface. A set threshold of correlation criterion is introduced to eliminate image mismatch, and the reference image is correspondingly updated to conduct later matching under large deformations of structures. A shaking table test of a 3-story frame model under El Centro earthquake is performed in lab. Experimental results indicate that the displacement curves acquired by the proposed stereovision method and linear variable differential transformers agree well, and this testifies the reliability and effectiveness of the presented stereovision method. Compared with our previous stereovision method, the proposed stereovision method has a higher precision and faster running speed. This demonstrates that the proposed stereovision method is more suitable for the actual use and can monitor the entire process of structural collapse.