Bridge Crack Detection Based On Uav Vision Technology Research

Intellectualization and automation are the inevitable trend of the future development of civil engineering industry.At present,the scale of China’s infrastructure construction is constantly expanding,and the number of bridge projects is also increasing accordingly.China’s total number of Bridges has jumped to the top in the world,with a total of over 800,000 at the end of 2017.However,under this challenge,the main method of bridge appearance detection in China still relies on the bridge inspection vehicle to observe manually.This method not only takes time and energy,affects traffic and safety risk,but also fails to guarantee the detection accuracy.These problems can be avoided by using uav for bridge appearance detection,and the development of recognition program for bridge crack image is of great significance for the automation of bridge appearance detection.In this paper,the characteristics of uav’s bridge image acquisition and the difficulties encountered in flying under the bridge are studied,and the flight mode of uav with fixed height and distance is proposed.Based on this,the hardware design and working principle of uav are proposed.According to the characteristics of the unmanned aerial vehicle(uav)to collect the image,in the process of image preprocessing,this paper mainly in view of the image gray processing,filtering noise reduction,edge detection,threshold segmentation,morphological operations such as doing the comparison and analysis,the threshold of edge detection based on artificial adjustment method and based on the selected threshold image segmentation method,the experimental results show that the practicability and reliability of the method.In the process of fracture image recognition and analysis,this paper mainly proposes a bridge fracture recognition method based on the connected domain marker features.A method of fracture parameter measurement based on image refinement and average width is proposed.Comparing different corner detection algorithms,Harris corner detection method based on manual selection threshold was proposed to ensure the timeliness and accuracy of detection results.Aiming at image fusion,a trial and error method is proposed to select the weighted coefficient,and the results show that the method can effectively remove the seam.Finally,an image processing software for bridge cracks was developed,and the reliability of the software processing results was verified by taking photos on the spot with uav.The three-dimensional reconstruction of the bridge model through sequence photos achieved the accurate description of the location information of cracks.Compared with traditional detection methods,bridge detection based on uav vision improves detection efficiency and accuracy,reduces detection cost and safety risk,and is of great significance for realizing intelligent and automatic bridge detection.