Research On Automatic Identification Of Steel Bundles Based On CCD/Structural Light

The three-dimensional scanning measurement technology based on monocular vision has the advantages of high measurement accuracy,fast dynamic response,high efficiency,high resolution,etc.In recent years,it has been widely used in many fields such as rapid product design,product quality inspection,safety control and object recognition.However,the quality inspection of steel coil packaging in most domestic steel enterprises is still in the state of manual inspection.Therefore,the research on automatic identification method of steel coil strap based on monocular vision has important technical value and broad application prospects.In this thesis,aiming at the low efficiency of manual quality inspection of steel coil packaging,an automatic identification method of steel coil strap based on CCD/structured light is proposed.The method is essentially based on monocular vision3 D scanning measurement technology,mainly through laser emitter,light source,CCD industrial camera,computer and other devices,build a three-dimensional scanning measurement system based on monocular vision,automatically identify the steel coil strap,complete the quality inspection of steel coil packaging.Compared with the traditional artificial quality detection method,the method has the characteristics of safety,high efficiency,high resolution and accurate measurement.The research work of this thesis mainly includes:1.For the automatic identification method of steel coil straps based on CCD /structured light used in this thesis,the research background and significance are elaborated.The research status of structural light measurement technology,structural light stripe center line extraction technology and vision-based 3D reconstruction technology are summarized.The three-dimensional scanning measurement technology based on monocular vision is used to automatically identify steel coils and complete the problems in the quality inspection of steel coil packaging.The research content and overall solution of this subject are proposed.2.For the distortion of the images collected by industrial cameras,the related algorithms of camera models and camera calibration are studied,and the CCD camera is calibrated by the classic calibration method invented by Zhang Zhengyou.The internal and external parameters of the camera are obtained,and then the distortion problem of the acquired image is corrected according to these parameters.3.Aiming at the problems of large calculation amount and slow operation speed of traditional structured light stripe centerline extraction algorithms,the advantages and disadvantages of various algorithms for structured light stripe centerline extraction are compared and analyzed.The refinement algorithm invented by Zhang-Suen is easy to operate.Steger algorithm has good robustness and high precision.The characteristics of these two algorithms are more prominent.Considering these two algorithms,the image is subjected to structural light stripe center line extraction.After comprehensive comparison,Steger algorithm is selected and the problem of large computational complexity is improved.Finally,the improved Steger algorithm is used to extract the structured light stripe center line of the image.The improved Steger algorithm not only has the characteristics of high precision and robustness of the Steger algorithm,it can process an image with a resolution of 761.8 milliseconds,which can meet the real-time and high-precision requirements of automatic identification of steel coil straps.4.According to the characteristics of steel coil strap detection in steel plants,the advantages and disadvantages of various algorithms for image feature extraction and matching are analyzed and compared,and the related algorithms are used in this thesis to test and analyze.Feature extraction and matching were performed on the images processed in the previous period,and several classical feature extraction and matching algorithms were compared and analyzed.Finally,the ORB algorithm is used to extract and match the image.The algorithm improves the detection speed while ensuring repeatable and anti-noise.It can be used for real-time system feature extraction.5.Based on the camera calibration algorithm,structural light stripe center line extraction algorithm,image feature extraction and matching algorithm,a simulation experiment platform for automatic identification and 3D reconstruction of steel bundles based on CCD/structured light is designed.Through simulation test,the reliability and real-time performance of related algorithms are verified.