Analysis And Research On The Particle Size Of Concrete Materials Based On CCD

With the rapid development of society,in real life,concrete is widely used in various fields such as road and bridge construction,building,decoration and so on.So it puts forward higher requirements to the property index of concrete,but there are many factors that may affect the property of concrete,such as the water consumption,water cement ratio,sediment concentration,cement type,aggregate condition and admixture of concrete.At present,due to the use of the past equipment and methods,the matching processes relying on experience completely rather than quantitating to control the matching process based on the requirements of the object.In this paper,CCD camera is used to obtain aggregate image before the concrete aggregate entering the mixing equipment.The combination of the adaptive fuzzy C mean clustering algorithm and marked watershed algorithm are used to segment the adhesive particles in the image with the MATLAB simulation platform and digital image processing technology,which can realize the functions of automatic segmentation,edge detection,particle marking,detection and statistical analysis of concrete aggregate image.The main jobs of this paper are as follows:Firstly,determine the research program used in this paper.Analyze particle size of concrete material by digital image processing technology.Taking the site environment and the requirements of this subject into account,the CCD camera is chosen as the tool to gather images and the calibration method is selected as the image calibration method.According to the analysis of image processing scheme,we determine the image segmentation method,at the same time,make the characteristic parameters of concrete granular material clear.Secondly,the collected particle images are preprocessed with the image processing techniques.Firstly,do gray image process.Then,compare and analyze the effect of mean filter,median filter and gradient inverse weighting method,from which the more appropriate method is selected.And then make the image equalizing to enhance image contrast.In the process of threshold binaryzation,Iterative threshold segmentation and Otsu method are used to segment the image,which can complete the binaryzation of image.Thirdly,in this paper,an adaptive fuzzy C mean clustering algorithm and a marked watershed algorithm are combined to segment the adhesive concrete material image,the method of determining the fuzzy weighted index mand the method of determining the number of cluster centers care also put forward.By making the foreground and background of the concrete material image,and do extreme value calibration of each region,the segmentation of adhesion particles is completed finally.Lastly,based on the camera model,the camera internal and external parameters are obtained by Zhang Zhengyou calibration method.Then the relationship between the camera coordinate system and the world coordinate system is used to convert in order to getting the proportion of the image and the actual value of particles.According to the result of segmentation,the characteristic parameters of rock particle size are extracted,and the results of the experiment are analyzed.Through the research of this project,the adhesion image segmentation of the rock aggregate concrete based on processing fuzzy C means clustering algorithm and marker based watershed algorithm is realized,which can obtain the aggregate particle size information in real time.If the method proposed in this paper is applied to the actual production,the efficiency and performance of concrete production would be greatly improved,meanwhile,the economic and social benefits of the users would be also improved.