Detection Method Of Wedge Based On Line Scanning Three-dimensional Point Cloud

Anchor wedges play a decisive role in the safety performance of important infrastructure such as High-Speed Railway,bridges and Five-hundred-meter Aperture Spherical Telescope.Anchor wedges have defects in the automated production process,but traditional detection methods cannot meet the requirements of on-line detection accuracy and efficiency,high-precision and high-efficiency detection is an urgent technical problem in the industry.Aiming at this problem,this paper proposes and studies the basic theory and method of online detection of clips based on line scan 3D point cloud.The research content of this dissertation are as follows:(1)A wedges detection method based on three-dimensional measurement of line scan is proposed.Based on the study of wedges structure and typical defect features and comparative analysis of existing three-dimensional detection methods,a three-dimensional line scan method is proposed.It has the advantages of high detection efficiency,high precision and simple structure.(2)A wedges line scan 3D inspection system is designed.According to the requirements of the actual detection technical specifications of anchor wedges,the line scan three-dimensional measuring sensor is designed is designed and an improved standard displacement calibration method is proposed.A three-dimensional detection system for wedge line scanning is designed.The system section data acquisition frequency is 1.6KHz,and the lateral range and resolution are 260 mm and 0.0635 mm respectively under the detection speed(including conveying and detection)? 0.45s/pair;the elevation range and sub-pixel resolution are 25 mm and 0.00198 mm respectively.(3)A three-dimensional measurement extraction and defect recognition algorithm based on point cloud is proposed.Based on the analysis of the data error source and characteristics,the data is preprocessed to obtain effective point cloud data.Studying the characteristics of point cloud data,combined with the parameter extraction requirements,put forward the data processing ideas of "section level" and "surface level".Design the point cloud data extraction from the thread area,improve the peak finding algorithm to locate the thread teeth,and reconstruct the quantitative thread based on the model.A three-dimensional parameter extraction algorithm for wedges composed of steps such as teeth.The wedges defect was identified based on the improved region growing algorithm,and the recognition accuracy reached 96.18%.(4)Experimental verification.Set up an experimental platform to collect 3D point cloud data on the inner surface of the wedge.Repeated experiments,calibration experiments,clip sample test experiments,statistical analysis of experimental results,system repeatability of less than 10%,the ratio of the difference between the crest and the base of the tooth and the fitted straight line after the calibration is reduced by an average of about 1.5.This thesis studies the wedges detection method based on line scan 3D point cloud,which solves the problem of high precision and high efficiency detection of clips.The test results can fully evaluate the clip quality.To meet the real-time testing requirements of the production line,it provides theoretical and experimental basis for the design of the anchor wedges online detection system.