Research On On-line Thickness Measurement Method Of Aluminum Ingot Based On Image Processing

With the continuous development of China’s economy,the demand for non-ferrous metals is gradually increasing,as a metal with light weight,corrosion resistance,low density,low melting point and good plasticity,the aluminum is favored by the manufacturing industry.However,due to the immature conditions of direct transportation of liquid aluminum and the worrying safety,it is usually carried out by long-distance transportation with aluminum ingots.Therefore,the thickness of the aluminum ingot is required to be high,and if it is too thick or too thin,the aluminum ingot form will occur.Changes will eventually result in the aluminum ingot not meeting the factory requirements and being directly scrapped.Therefore,real-time measurement of the thickness of aluminum ingots in the manufacturing process of aluminum ingots has practical and practical value.At present,the thickness control of the aluminum ingot mainly relies on manual visual observation and ensures the thickness of the aluminum ingot to meet the production needs by controlling the amount of aluminum liquid.This method of control has problems such as low precision and adjustment lag.There is a method for improving the use of laser time-of-flight ranging.However,problems such as thermal radiation in the molten state and surface turbulence of the aluminum liquid also affect the accuracy of the measurement,and the cost of the method is relatively high,which affects the economy.In view of the above problems,this paper carried out research on the online measurement system of aluminum ingots.By analyzing the aluminum ingot manufacturing mold,the thickness of the aluminum ingot was converted into the surface width of the aluminum ingot,thereby measuring the thickness of the aluminum ingot by a non-contact method.At the same time,the aluminum ingot image is photographed by the monocular vision measurement method,through the conversion of color model and the extraction of color space components,the useless information outside the aluminum ingot is removed.Comparing the advantages and disadvantages of various edge detection operators combined with the correlation algorithm of threshold segmentation,the information of the aluminum ingot part is improved,and the image of the photographed aluminum ingot is optimized by image morphology operation.Finally,the thickness information of the aluminum ingot is obtained by the Hough transform method.Through the simulation of the whole measurement system,the system for real-time measurement of aluminum ingot thickness is compared with the widely used artificial thickness measurement and improved laser time-of-flight ranging method.The proposed method is used in measurement accuracy and cost optimization.There is a big improvement,which will provide a great help for the final investment in production.