Research On Image Processing Of Concrete Cracks Based On LAB And OTSU Algorithms

At present,concrete works have been widely used in river,lake,water conservancy,housing construction,national trunk highway and railway subgrade slope management.Cracks in concrete engineering is a kind of common phenomenon,it produces,not only can reduce the permeability resistance of concrete,concrete external perception experience,will directly affect the use of concrete.The production of cracks will make the reinforcement corroded and concrete carbonized,and will greatly reduce the durability of the material and reduce the service life of concrete.The distribution of cracks is very complicated,and it has great influence on the structure.Therefore,the study of cracks is of great significance.Aiming at the huge problem of manual measurement,the computer can get the result easily and quickly in image processing.Pixel detection technology based on image processing has been widely used in engineering and has obtained many research results.Currently using support vector machine(SVM),artificial neural network,such as pattern algorithm of image edge detection of cracks,there is a lack of measurement precision was not well,parameter selection,parameter extraction accuracy is high enough and need the problem such as long processing time.Aiming at the concrete crack detection of low efficiency and low accuracy of classification,proposed the LAB algorithm and OTSU algorithm with the combination of image processing technology to improve the effect of image edge detection and reduce the interference of impurity,reduce false the impurity into fuzzy situation of edge pixels,increase the accuracy of the feature parameter extraction,reduce the running time,better for computer image measurement technology is applied to concrete engineering provides a effective means of detecting cracks.The main contents of this study are as follows:1.Using image segmentation technology,a preliminary grayscale processing is performed on the collected images.In the process of processing,the space complexity is reduced by using the second improved segmentation method,and the time consumption is reduced.According to the characteristics of the collected concrete crack samples,this paper introduces the method to filter the concrete crack image by means of the mean filtering and other image denoising methods to reduce the interference of impurities in the image.Theedge detection algorithm of a variety of classical operators is analyzed to detect the edge of concrete images and extract the target image.Put forward based on LAB and OTSU algorithm,the combination of color domain widely using LAB color space model and the characteristics of the OTSU algorithm is simple and practical,in order to improve the previous in image denoising,edge detection,the effect is not obvious.Through this method,not only can reduce the time of consumption,and can be targeted to remove noise and enhance the effect of edge detection,can under the condition of abundant color gamut better separation of image background and target,not discrete integral problems.It lays a solid foundation for the extraction of characteristic parameters of concrete cracks.2,in the LAB and OTSU algorithm that combines image processing,(1)on the basis of extracting the joints and cracks of concrete crack area,(2)joint fracture area/total area,(3)joint crack branch number per unit area,(4)branch number,(5)the average size/sampling surface area,(6)the total length,(7)the average width of the seven characteristic parameters,through more comprehensive parameters selection,to increase the complexity of cracks.And then compared with the concrete crack parameter values measured in the field.In this way,the influence of each parameter on the evaluation result can be considered more comprehensively,and the error size between the characteristic parameter values and the measured values measured by the computer can be obtained quantitatively.The practical application of this method in the extraction of concrete fracture parameters is verified by the example of the crack image of complex concrete.