Research And Implementation Of Spring Measurement System Based On Machine Vision

At present,the measurement of spring and other parts mainly adopts manual measurement,which has low efficiency,poor accuracy,low stability,and the measurement data cannot be stored in real time,so the processing speed and accuracy cannot meet the requirements.Machine vision is a new technology which combines vision and image processing.It can achieve the integration of optical communication.Most of them are applied in intelligent devices such as industrial robots.The development prospect is very good.In view of the requirement of high precision two-dimensional dimension measurement in industry,a spring dimension measurement system based on machine vision is designed by using the function of image processing software.The hardware structure and software processing of the system are studied.The key technologies of the machine vision system are expounded.The main contents are as follows:1.Spring image acquisition.The cylindrical spring being measured is placed on the platform.The height image of the spring is acquired at the clearest position by camera and double telecentric lens with parallel light source.The circular image of the spring is acquired by double telecentric vision inspection station in "Industrial 4.0 Intelligent Factory",which is outer circle and inner circle image.2.Preprocessing spring image with machine vision software HALCON.Firstly,the collected spring image is grayed,and then the gray image is sharpened by spatial filtering,which effectively removes the noise in the image,makes the edge contour clearer and then divides the inner circle and outer circle of the spring for laying the foundation for selecting the interested region.3.Canny operator and bilinear interpolation algorithm are combined to select the region of interest of the image and extract sub-pixel edges.In the process of locating the spring position,the shape-based template matching method is used.The selected spring contour area is used as the template of shape matching,and then the image pyramid is used to search the position of the template quickly.Affine transformation is applied to the template image to make the matching result more accurate.4.Fitting the inner and outer circles of springs with the least square method based on Tukey to get the dimensions of inner and outer diameters.According to the minimum outer rectangle generated by the height image of springs,the outer diameter and height dimensions can be obtained.Finally,the measurement work of converting the pixel level into the actual millimeter level can be completed by system calibration.Then,errors of the dimension measured by the HALCON software designed by ourselves and the actual size of spring is analyzed.5.Use VisionBank visual inspection software to measure the inner and outer diameters of springs,and compare the measured results with the results measured by HALCON software designed by ourselves and analyze the errors.