| As an important dielectric ceramic category,talc ceramic materials have abundant material-related resource reserves,dielectric loss advantages,and low cost.Therefore,its talc porcelain parts play a key role in insulation protection in many electronic products,and the market demand is large.However,in view of the fact that talc porcelain parts usually have problems such as large batches,various kind batches,complex size specifications,and unstable manual implementation standards.Production units have certain difficulties in inspecting the dimensional quality of parts.Based on the basic requirements of talc porcelain parts inspection and research,this paper proposes a non-contact machine vision parts dimension measurement system.The main contents are as follows:(1)Based on the dimensional acceptance requirements of talc porcelain parts,we start with the digital image acquisition module on the surface of the parts.The hardware of the module platform is selected and determined to suit the needs of image acquisition and image processing.The mono-mode industrial camera is used to directly obtain single-channel images with sharper edges,which saves the computing power required for data dimensionality reduction,and uses top-hat filtering and high-pass filtering image edge enhancement methods to enhance image global contrast and graphic edges for grayscale images sharpening.(2)Through threshold segmentation,the surface image of the part is expressed as a binary image of the digital image,and then the target graphic area of the image is highlighted and the redundant graphic information is weakened.In view of the size measurement requirements of image information,this paper performs multiple morphological opening and closing operations of specific convolution kernels on the image in the state of image binarization to optimize the edge continuity of the image and reduce the interference of pixel particles that are not the characteristics of the image itself.The part surface geometry involved in the controlled dimensioning of the part is considered.In this paper,the contour seeking method based on geometric topology relationship is used to extract and store the inner and outer contours of the part surface effectively.The rectangular and circular geometric parameters and pixel coordinates of the surface of the part are extracted by the graphic detection method based on the Hough change principle,and the control size of the part is obtained through the calibrated pixel conversion ratio.Using the target part as the verification sample,the deviation analysis of algorithm and gage measurement is carried out.(3)The whole machine is set up for the machine vision dimension measurement system of talc porcelain parts.The visual measurement system includes a feeding and transmission part,a machine vision inspection part,a blanking and muti-class accepted product screening part,and a terminal software operation control part developed based on the Qt platform.Through prototype development,on-site part quantity testing and machine debugging,this machine vision dimensional measurement system can effectively monitor the size and quality of talc porcelain steadily over time and perform multilevel classification of good parts according to the specified specifications.The equipment has passed the acceptance inspection in the parts production unit,has reached the production and use standards,and has certain engineering practical value. |
