Research On Image Template Matching Algorithm Of Scribing Machine Based On Circular Projection

With widespread use of scribing machines in industrial applications,the requirements for automation,high efficiency,and high precision are becoming higher and higher.The visual positioning system of scribing machines is a prerequisite for accurating recognition and cutting workpieces.With the LED package board image as the matching target,the visual positioning system of the automatic scribing machine is designed with high precision and good stability.Mainly include the following:First of all,according to the requirements of the visual positioning of the scribing machine,the software framework of the visual positioning system of the scribing machine is designed.Preprocessing the image collected by the camera,using gray transformation,image filtering,morphological transformation,image sharpening and other methods to improve the quality of the image and prepare for template matching.Then,according to the special shape of the LED package board image,the process of obtaining the template matching algorithm parameters is designed.After extracting the centroid of the target contour,the corner of the template is determined using corner detection and geometric relationships.According to the principle of template matching algorithm,the radius of the template is modified to obtain the parameters required for template matching.Aiming at the problem that the traditional template matching algorithm is complicated in calculation,poor in robustness and cannot be accurately matched when the template image is rotated,the circular projection template matching algorithm with rotation invariance is designed.Comparing the similarity of the circular projection vectors between the template image and the search subgraph to determine the matching position,and the layering strategy based on the image pyramid is used for acceleration.Aiming at the problems that the classic circular projection vector is sensitive to noise,grayscale change,and contrast change,a robust circular projection vector is reconstructed.To solve the problem of large amount of calculation in the circular projection template matching algorithm,the transform of the circular projection vector sum based on Fourier transform is used to accelerate.Finally,the performance of the algorithm is tested using the LED package board image.The results show that the circular projection template matching algorithm can be accurately positioned under any rotation angle and can work normally under the influence of noise,contrast change,grayscale change and other interference.