基于机器视觉的微细孔壁缺陷检测关键技术研究-测试计量技术及仪器专业论文.docxVIP

ABSTRACT Small size and thin features of pipes or boresare used more and more widely in the aerospace, automotive, energy and chemical industry.It is very important to achieve the defect detection and measurement of thin pore wall morphology for modern industrial manufacturing, quality and safety controlling. Among current methods, such as endoscopes, optical probes and micro-pipe robot, there arealways deficiencies such as needed for manual intervention for sampling, high structural complexity or low efficiency, which ishard to meet the actual requirements. In this paper, a way to achieve a flexible micro-pore wall defects, automatic, rapid, high-precision measurement method is researched. With the use of special optical transmission components, light source importing and image exportingis completed. Combined with the principle of machine vision and image processing algorithms,pores on the inner defect size and location of the high-precision detectionis achieved. Through modeling and experimental analysis,system performance evaluation is completed. The main work of this paper is as following: Existing defect detection method inner pores surface is analysd.With the objective needs of industrial production and quality control, a flexible, fast, comprehensive, high-precision measurement method based on machine vision is researched and applied to thin pore defect detection. The method of attitude adjustment of optical transmission component is researched. By constructing a mathematical model of image transmission and feedback control based on image processing, the alignment between optical transmission component and the measured pores is completed to ensure accuracy and safety of detection system.Experimentsare done to verified the feasibility of attitude adjustment. Adaptive illuminator systems based on imageevaluation is researched. MCU is used on high-power white LED light source for embedded control to ensure the original image clarity and stabilityunder different