Research On Image Processing Algorithm Of Laser Marking QR Code On Aluminum Ingot Surface

Laser marking is an important method of direct part identification technology because of its non-contact,high speed,and low cost.It has become an important choice for material identification,and is widely used in metal,glass,ceramics,and plastics.Aluminum is an important industrial raw material and mainly provided to downstream customers in the form of cast aluminum ingots.According to national standards,the surface of aluminum ingots must have two-dimensional code labels,and it is of great value to apply laser marking technology directly to aluminum ingot product identification to reduce cost and improve efficiency.There are difficulties in the application of laser marking on the surface marking of aluminum ingots.The main reason is that the surface of the aluminum ingot is rough and bright reflective,which leads to the unclear labeling of the laser marking part,the lower quality of the two-dimensional code image in the label,and serious light pollution.The background noise is high and the contrast ratio is low,which makes it difficult for the two-dimensional code to be directly recognized by common equipment.Therefore,it is of great significance to study the image processing algorithm of laser marking on the surface of aluminum ingot and improve the recognition rate of two-dimensional code,and the laser marking technology for the electrolytic aluminum industry has been promoted.This thesis first analyzes the laser marking of two-dimensional code images on the surface of aluminum ingots.Based on the common image binarization method and image segmentation method,the aluminum ingot barcode image is processed.The results show that: 1)The global or local threshold method has no obvious effect due to the low contrast of the image.2)Image segmentation method based on color,texture and edge extraction is affected by light pollution and background noise,and the result is not ideal.3)The background noise is similar to the two-dimensional code area and the processed image is still unrecognizable.Aiming at the above problems,this thesis proposes an image processing algorithm for aluminum ingots,which firstly improves the effectiveness of color features by transforming the color space,and enhances the image texture based on Gaussian difference.Based on this,the color features and texture features are performed.Fusion clustering,image segmentation,morphological optimization of theoutput of two-dimensional code image to complete the decoding calculations.A large number of two-dimensional code processing experiments prove the effectiveness of the proposed image processing algorithm.Finally,based on this algorithm,a laser labeling aluminum ingot inbound storage management system was developed,and based on the Android mobile platform,the inventories of laser marking aluminum ingots were implemented,which improved the product management efficiency of aluminum ingots.