Video Detection Of Bridge Cable Surface Defects Based On MATLAB

China's terrain is heading west high and low east,and the mountain area accounts for more than two-thirds of the total land area of the country.Therefore,large-span cable-stayed bridges and suspension bridges are widely used,and the number of cable-stayed bridges in China accounts for more than half of the total number of cable-stayed bridges in the world.In recent years,the development focus of China's bridge industry has gradually shifted from "maintenance-oriented" to "maintenance-oriented".At present,among the cable-type bridges that have been built in China,the cable is the main bearing member,and the health status of the cable directly affects the normal operation of the bridge.The defect detection of the cable appearance is one of the key contents in the bridge safety inspection work,and it is also an important content to discover the safety hazard of the bridge cable.However,in the actual work in the industry,the surface detection of the cable is still carried out by means of manual visual video data.The drawbacks of this method are mainly low work efficiency;relying on personal subjective experience to judge;long-term visual inspection is prone to errors.Therefore,based on the actual needs,this paper develops a video detection software system for bridge cable appearance,and carries out technical research on key problems in the system.Experiments show that the system can replace the manual defect detection of bridge cable appearance.This thesis focuses on the theory and technical process of the appearance defect detection of cable-type bridges.Relevant experiments were carried out around the actual engineering requirements for the automatic detection of video data of cable-type bridges.A reasonable compression criterion is adopted for the original video data to extract the suspicious image,which reduces the amount of data.At the same time,the threshold segmentation precision of the defect image is improved by the introduction of morphology.Then the segmentation image is extracted and the defect geometry information is used to quantize the target region.Finally,the target region is developed.The corresponding video processing software.Experiments show that the software developed in this paper has a certain role in the appearance detection of cable-type bridges,and its technical processing flow also has certain application value in video detection in other fields.The main contents and results of this paper are as follows:1.Research on video compression methods based on block and statistical features.The original data in this paper uses the crawling robot as the carrier to obtain the video data of the cable surface.The amount of data is huge.This paper proposes to obtain a single frame image through video decoding,and then obtain the video quality,moving speed and field of view according to external devices such as camera.The range size and other parameters set the sampling interval to further extract the image compression data.Finally,by segmenting the image,the suspected disease image is extracted according to the statistical features of the image sub-block,and the data amount is compressed without missing the disease image.2.Research on Threshold Segmentation Method Based on Mathematical Morphology.According to the bridge cable image obtained in engineering practice,this paper analyzes and verifies the defects and shortcomings of the general threshold segmentation method.Through the analysis of a large number of cable defect images,due to the complexity of the background and the noise of the noise,the mathematical morphology operation is introduced in this paper,which can better preserve the edge while removing the image noise.Experiments show that the segmentation effect is significantly improved after structural element processing,and the positioning is more accurate.3.The target geometric feature is obtained by segmenting the image connected domain marker.The connected domain mark is performed on the segmented image,and different regions are marked with different numbers.The region is tracked by different connected domain codes,and geometric information such as image area,perimeter,and circumscribed rectangle length and width of different defect regions are separately calculated to quantitatively describe the image features.4.Development of video detection software for bridge cable appearance defects.In this paper,after designing the technical process of video data processing for bridge cable appearance detection,the software interface and corresponding data processing functions are designed through MATLAB platform,and the video data of a cable-stayed bridge crawling robot is tested and verified,which proves the proposed in this paper.The correctness of the technical process and the applicability of the software.