Research On Measurement Technology Of Large Size Parts Base On Machine Vision

With the speeding up of the manufacturing and production,high requirements are proposed to quality,accuracy and speed of part detection.The traditional contact measurement method is limited in measurement time and subjective factors.It will also seriously impact on the efficiency of product detection.In recent years,the machine vision measurement gets up and running for the advantages of non-contact,high speed acquisition,high precision,real-time display.It has become an effective solution to solve the problem of parts gauging in the manufacturing industry.Based on machine vision related knowledge,non-contact on-line measurement for large parts were studied,mainly around the related machine vision image preprocessing,image matching,image edge detection and geometric measurement research.First of all,the development history of measurement by the use of machine vision and the significance of the research was introducted.The classification of measurement technology and the research status at home and abroad are summarized.Secondly,the study of the related content based on machine vision measurement.In the study of image preprocessing,the image gray processing and image denoising algorithm are analyzed.At the same time,the classical algorithm of image denoising is also analyzed and compared.The results prove that the wavelet denoising algorithm is suitable for this project.In the study of image mosaic algorithm,a SIFT(scale invariant feature transform)matching algorithm based on feature modification is adopted on the basis of the current image mosaic algorithm.The validity and superiority of image mosaic are verified by experiments.In the study of edge detection and measurement principle,the classical edge detection algorithm is analyzed and an improved single pixel edge detection algorithm based on Canny edge detection is adopted.On the basis of this,the sub pixel edge detection is further studied and the detection algorithm is verified to extract the edge of image accurately.In addition,the measurement principle and system calibration method are studied in the measurement.Finally,through the above algorithm analysis,a large scale parts measurement system is developed based on machine vision in Visual C++ 6.0,and a lot of experiments are carried out.The experimental results can be verified,the measurement system can be measured precisely and quickly within the error range,and achieve thedesired purpose.