Method for automatically detecting and recognizing bridge cracks based on computer vision

Abstract

The invention discloses a method for automatically detecting and recognizing bridge cracks based on computer vision. By constructing a training depth network model, taking a captured image as an input, and calculating respective hidden layers, the method finally outputs the classification label of the image to realize crack identification and complete the computer's understanding of the input image content. The method, for the automatic monitoring and recognition of bridge cracks, and realizes the whole process automatic processing of model training, crack recognition and result display of crack images of real steel box beams containing complex background interference information. The method is convenient and accurate, and improves the efficiency of bridge crack detection and the accuracyand stability of the detection results.

Description

technical field [0001] The invention relates to the field of civil engineering monitoring, in particular to a computer vision-based automatic monitoring and identification method for bridge cracks. Background technique [0002] With the rapid development of my country's national economic construction, more and more large-scale infrastructure construction plays an extremely important role, especially large steel box girder cross-sea bridges. Due to the long-term complex vehicle loads of large steel box girder cross-sea bridges, the welds of steel box girders often have different degrees of fatigue damage accumulation due to the existence of initial defects, and then fatigue cracks are formed. Under the coupling action of disaster factors such as the long-term effect of load, fatigue effect and mutation effect, fatigue cracks will expand along the direction of the weld or to the roof, diaphragm and other components, resulting in attenuation of the resistance of the bridge stru...

AI Technical Summary

Problems solved by technology

In the process of traditional image processing, these artificial marks and handwriting have brought great interference to the recognition of real cracks

And the size of fatigue cracks is relatively small, and some crack widths are only 10 -1 mm level, it is easier to be treated as noise in the traditional image processing process

In addition, some identification methods also require the internal and external parameters of the camera for image capture (such as object distance, image distance, shooting angle, etc.), or require additional professional measurement equipment

Overall, traditional fracture identification methods require too much manual intervention and are expensive

Images