| Under the guidance of "made in China 2025" strategy,a large number of domestic shoe-making enterprises and scientific research institutions have been committed to the transformation and upgrading of automatic production technologies of shoe-making industry,and have achieved good results so far.At present,mast of the achievements are to transfer the fixed shoe mold informatipn to the industrial robot,and then control the robot to realize the polishing,glue spraying.and other proces s operations of the sole.However,the robot can't obtain,external information in real time in this way,which makes the robot ignore th initialpositioning deviation in the process of sole processing and the possible pssition change in the transmission process when it is grinding and glue spraying process,which reduces theyield of footwear and affects the production efficiency and efficiency of the enterprise.Aim ing at solving these problems,the related technologies of the three-dimensional data recovery of the sole in the polishing process of the sole in the shoemaking industry based on binocular vision were studied,mainly include image acquisition,camera calibration,image processing,stereo matching and three-dimensional reconstruction.Furthermore,the three-dimensional data of the sole was calculated by building a binocularvision platform by the theoretical analysis.The main work of this paper includes the following aspects:(1)The camera imaging.model and the three-dimensional measurement principle of binocular vision were analyzed to obtain the mapping relationship between the three-dimensional information of spatial point and the pixel points of two-dimiensional plane image.The difference between the conventional binocular model and the ideal model was explored,and gain the key parameters affecting the measurement accuracy of the binocular vision measurement system.(2)The binocular vision system experiment platform was built and calibrated.The hardware selection and construction of the experimental platform are realized combined with the project requireiments and the analysis of point(1).The principle and method of ZHANG Zheng You's calibration algorithm applying in the system were explored.On this basis,the binocular vision system was calibrated by using MATLAB and OpenCV to obtain the internal and external parameters,distortion parameters and the position and attitude relationship between the two cameras,so as to achieve the corrected image.(3)The fundamental theory of image preprocessing and stereo match algorithm were analyzed.The common image denoising,histogram equalization,threshold segmentation,Canny operator and morphological processing algorithm were analyzed,so as to eliminate the image noise produced in the process of image shooting,transmission and preservation and reduce the interference items in stereo matching.Furthermore,the algorithms were implemented based on OpenCV,and the results weee briefly analyzed.Then,stereo matching algorithm was Introduced from the aspects of constraint principle,.classification and implementation steps based on Schrstein and Szeliski.Finally,SGBM algorithm in Open CV was analyzed and disparitymap was obtained by using this algorithm.(4)Recoverying,three-dimensional information of the sole and experimental research:the method of recoverying three-dimensional information of the sole based on disparitymop was studied,so as to calculator three-dimensional coordinates in the world coordinates system by using disparitmap and reprojection matrix On this basis,tree-dimensional reconstruction software based on binocular vision which applying in footwear flexible body was compiled by using MFC,OpenCV and PCL.The software realizes the functions of image acquisition,camera calibration,image processing,stereo matching and three-dimensional reconstruction which verified the feasibility of the application of binocular vision technology in the footwear industry.It had certain guiding significance for applying binocular stereo vision technology in the shoe-making industry. |
PDF Full Text Request