Solar Cell Sorting System Based On Machine Vision

With the development of society,environmental problems and energy problems have gradually become prominent,solar energy as a clean energy has become one of the important directions of the development of new energy,solar cell as a carrier of power generation has been widely used.Under the influence of external factors or production process,the solar cell will produce defects such as missing angle,dirt,scratch,broken grid,color difference and so on.These defects affect the power generation efficiency,service life and overall beauty of the solar cell module,so it is necessary to separate the cell.At present,manual detection is the main detection method,which is affected by subjective factors,and the detection efficiency and accuracy are not high.In this paper,a solar cell sorting system based on machine vision is designed by combining machine vision technology and image processing technology.The specific content of this paper is as follows:(1)Complete the design of the overall scheme of the system and build the hardware platform of the system.The selection of camera,lens,lighting unit and the design of camera obscura,communication module and motion control module ensure that the system can obtain clear images of solar cells on the production line.(2)Complete the pre-processing of the cell image.Firstly,the calibration plate is used to correct the distortion of the solar cell image.Secondly,the combined filter and gray linear transformation are used to enhance the image,highlighting the details of the target.Finally,Canny edge detection algorithm and affine transformation are used to segment and locate the ROI.(3)Design the defect detection algorithm of solar cells.Aiming at structural defects,a feature extraction algorithm suitable for angle missing,edge collapse and hole was proposed from the perspective of morphology and BLOB analysis.It could detect all structural defects simultaneously and retain the original convex shape of defects.Aiming at the dirty defects,the algorithm design is carried out from the aspects of gray morphology and improved threshold segmentation,which can overcome the problem that the gray level of the defect area is similar to that of the cell area,so that the dirty area can be extracted completely.Aiming at the defects of the gate line,the algorithm is designed from the aspects of Gaussian mask convolution and fitting to extract the parameters of the Taylor quadratic polynomial in the x and y directions of each point,which can completely extract the gate line region.(4)Design the color difference sorting algorithm of solar cells.Based on the idea of partition comparison,the cell region was divided into 25 equal parts,and the difference of maximum value,variance and mean value between regions were compared to determine the color difference within the chip.A new concept of color difference value was proposed for the interchip color difference,and the cell color difference sample library was established and the tap position was automatically divided,then the color was sorted according to the color difference value of the target cell.After completing the above study,test the system.The sorting results show that the average accuracy of defect detection is 97.48%,and the accuracy of color difference sorting is98.7%.In the system,the image processing time is less than 500 ms.The system designed in this paper has realized the function of defect detection and color difference sorting on the premise of high efficiency and high accuracy,and is now being productized and marketed.