Research On Automatic Location And Recognition Method Of DIE Chip Based On Machine Vision

With the increasing application of machine vision control technology in the intelligent manufacturing industry,a large number of intelligent manufacturing equipment based on machine vision control system technology has been developed.In the research and development of this kind of intelligent manufacturing equipment,there are many chip positioning and identification work.Machine vision technology cooperates with electrical control system to complete positioning,identification,detection and sorting operations.In this paper,aiming at the production needs of the actual intelligent device development,the automatic positioning and recognition method of die chip based on machine vision is deeply studied,and the method of die chip positioning and recognition is designed to meet the needs of the actual equipment operation.The main innovations of this paper are as follows:(1)In order to solve the problem of chip surface pollution caused by mobile chip in die chip detection equipment,an iterative minimum circumscribed circle positioning method is proposed.This method can effectively adapt to the edge contour missing and bulge caused by pollution.In each iteration,the minimum circumscribed circle of the edge contour is searched,and the number of edge contour points on the minimum circumscribed circle is taken as the weight value of the iteration,and the edge contour points on the minimum circumscribed circle passing through the iteration are deleted from the set of edge contour points.This method can effectively reduce the number of iterations and improve the accuracy of weighted average center.When the number of elements in the set of edge contour points decreases to a certain degree,the iteration ends.(2)Aiming at the problem of low accuracy and robustness of chip location,a feature location method based on image segmentation is proposed.This method combines the robustness of convolution neural network and the accuracy of digital image processing,uses image segmentation NeuralNetwork Model to segment the features used in image location,and then uses edge detection and morphological processing in digital image processing to accurately locate the features.The results obtained by this method are more accurate than those obtained by single convolution neural network method or single digital image processing method.(3)Aiming at the low efficiency of OCR character recognition in specific imaging environment,a method of OCR character recognition based on similarity matching is proposed.Firstly,the standard character template is calculated for a large number of the same characters,and the template is calculated by the character similarity of these standard templates.In the stage of character recognition,the similarity between the character to be recognized and the standard character template is calculated,and the character represented by the standard character template with the highest similarity is taken as the result of the character to be recognized.According to the positioning and recognition tasks of die chip detection station,the tasks are divided into coarse positioning,fine positioning,high temperature positioning and low temperature positioning,and OCR character recognition.From the point of view of digital image processing,a positioning algorithm is designed to meet the requirements of intelligent manufacturing equipment for coarse positioning and fine positioning.For high-temperature positioning and low-temperature positioning,digital image processing,U-Net segmentation positioning and SSD positioning methods are used to study the accuracy,efficiency and stability of positioning.For OCR character recognition,similarity matching method and EfficientNet classification recognition method are used.