Research And Implement Of Embedded Mold Protection Image Processing Technology

In recent years,various plastic products are widely used not only in the major industrial production,in daily life can be also seen everywhere.This drives the demand for plastic products in all walks of life.Therefore,more and more attention has been paid to the processing technology and technics of plastic products.Due to injection molding can produce and manufacture various products with complex shapes,and easily combined with computer technology to realize automated production,occupies an extremely important position in plastics processing.Injection molding technology relies on the mechanical equipment named injection molding machine.After injecting a product by the injection mold machine,It's time to open the mold and crowl out of molded products.However,in the working process,sometimes material remains in the mold chamber.This kind of phenomenon is likely to cause the damage to the mold,and more seriously,it may influence the quality of following products.Since the clamping force of injection molding machine in closing the mold are usually very large,in such a pressure,even far less than the hardness of the object will damage the mold seriously.Therefore,in order to protect the mold from damage and improve production efficiency of injection molding machines,it is of great significance to research and look for an effective method to detect whether there remains the residue in the mold of injecting machine.(1)We are in reference of a large number of relevant domestic and foreign literature,then make an physical investigation for the working process of injection molding,workshop environment and so on.According to relevant theoretical knowledge of mold protection and the actual situation of the workshop,first summarized the general requirements of mold protection algorithm,then using several different algorithms to detect residue according to the size,shape of the mold and the region residues probably occurring on.(2)By comparing with other mold protection algorithm,considering the actual work of the injection molding machine,the detection for the cavity need strong real-time,high accuracy.Therefore,this paper proposes a new algorithm based on wavelet passivation and support vector machine to detect image of cavity.At the same time,the embedded mold protection system is designed to verify the residue detectionalgorithm.(3)Residue detection algorithm in this paper firstly select the appropriate wavelet function for image passivation in order to highlight the mold cavity images,without template matching and image correction,which reduces the error caused by the offset,meet the characteristics of small memory capacity of the embedded system;Then store the values of pixels through in arrays,in the training mode of the algorithm,the data is sent to the support vector machine to form the residue judge mathematical model,in the running mode of algorithm,the judgment model is used to judge whether there is any residue in the cavity image.Support Vector Machine(SVM)is introduced in mold protection algorithm,which improves accuracy and real-time.We collected several sets of different cavity images offline in mold workshop.Then We program on PC to verify algorithm proposed by this paper combined with MATLAB and Visual C++.The experimental data show that the average detection accuracy for cavity image is above 90%,and average detection time is 0.914 s,which is sufficient to meet the work requirements of injection molding machine.