Research On Quality Assessment And Evaluation Technique Of2D-barcodes Using Laser Direct Part Marking

With the development of manufacturing information engineering, Producing the product marks which can be read by the machine vision automatically become an urgent thing, present marking methods work at random, which lead to poor quality, or low recognition ratio. As laser direct part marking technology is fast, automotive, permanent, and suitable for high-heat, wet, rusty and harsh industrial environments, in this paper, it is chosen to mark2D-barcodes, and considering the lack of technological guidance and quality control standards of the technology, the contents of this study are as follows, Laser marking test is designed based on single variable control method to analysis the impact of efficient vector step, effective vector’s time delay between steps, Q switch frequency, electric current, marking depth and surface roughness on marking barcodes’ quality. Technical parameter optimization of Laser Direct Part Marking based on orthogonal experiment was presented to find the best choices of process parameters, and also the test results proves that the efficient vector step is the significant factor marking barcodes’quality.A BP network model is established to predict marking barcode’s quality. Use the training samples to train the BP network, the simulation results show that the network has high accuracy in predicting the contrast. It proves that the prediction model is feasible and effective in theory and practice. Draw curves which describe the efficient vector step, Q switch frequency and electric current’s varies with contrast. It can be used as the processing parameters selection reference, which has important significance in getting high quality marking barcodes.Image acquisition system based on machine vision is designed, and its hardware and software are introduced. The low-angle LED lighting is the best illumination scheme after analyzing the impact of illumination on image acquisition. Image preprocessing algorithm is designed for laser marking barcodes on aluminum alloy. First of all, convert the image to binary image using OTSU. Then, use mathematical morphological method for image filtering. Third, use canny operator for image edge detection. The last but not one, initially locates the barcodes area with the help of horizontal and vertical projection. Finally, accurately locate the barcodes area using Hough transform. A large number test results proves that the algorithm is validity.An image quality evaluation which is suitable for laser marking2D-barcodes is designed. Images used for quality evaluation use the same image acquisition system and preprocessing algorithm. According to the difference of materials’shape, different quality evaluation index is set. When the material is flat, contrast, distortion angle, print growth and grid nonuniformity are set as quality evaluation index. When the material is curved, we need to consider the impact of different ratio of barcode’s width and material’s curvature on quality evaluation additionally. Divide image quality grades according to the scores of each index. The identification time of different grades barcodes proves that the image quality evaluation method is feasible. The method makes contribution to the foundation for the establishment of quality criteria system especially for laser direct part marking2D-barcodes.This work is supported by Independent Innovation Foundation of Shandong University (2010TS036).