Workpiece Contour Extraction Method In Machine Vision Measurement

The dimensional measurement of workpiece is one of the key links in machining industrial production.The accuracy of the measurement results will directly affect the final product quality and play a decisive role in the subsequent reprocessing and assembly.With the advance of the "Industry 4.0" era,the workpiece dimensional measurement technology is also developing rapidly,and people put forward the requirements of high precision,high efficiency,and automation for measurement.The traditional manual measurement method is obviously not suitable for today's mass industrial production environment.If the machine vision technology is applied in the field of industrial measurement,it can greatly improve the measurement precision,efficiency,reliability,as well as the level of automation and intelligence.Therefore,focusing on machine vision and workpiece dimensional measurement,we carry out systematic research and algorithm design in this paper.Firstly,illumination and image acquisition are analyzed and studied in this paper,a workpiece imaging and measurement system of blue LED annular lighting,5 million industrial CCD and double telecentric lens is designed.It lays a foundation for the acquisition of high contrast and high definition images.Secondly,we analyze the influence of various machining methods on the workpiece surface integrity,and further explain the influence of the workpiece surface integrity on contour extraction.After classifying the workpiece into "cliff" and "gentle slope" by its contour type,four corresponding contour extraction methods are proposed in this paper,including segmental extraction method based on template matching,upper surface information method,skeleton pruning method and gray histogram segmentation method.Some appropriate image processing algorithms are adopted in these methods to avoid the interference of workpiece surface integrity to contour extraction.Finally,the workpiece measurement system based on machine vision and the contour extraction methods proposed in this paper are used to measure the dimension of microwave resonator parts.The precision and accuracy of the contour extraction methods in this paper are evaluated through multiple sets of repetitive measurement experiments.Experiments show that these methods have strong practicability,since they can effectively eliminate the interference of special noise to obtain complete contour.Moreover,they have a good adaptability to different individuals of the same type workpiece,the detection results are not affected by the different gray scale of different individuals.Besides,they have high stability and precision in measurement,the results of repeated measurements evaluated by measurement uncertainty can be stable within the range of ±78).After measuring the standard gauge block,the absolute value of the measurement error is less than 138).The measured value is close to the real value,indicating that the contour extraction methods in this paper have high accuracy.