Research On Image Processing In Surface Defect Detection Of Solar Cell

With the rapid development of economic globalization,the in-depth development and extensive use of all kinds of new energy has attracted close attention from all countries in the world.Solar energy has the advantages of abundant resources and no pollution,and is one of the most renewable energy with scientific research value.However,during the manufacturing process or use of solar cells,the surface may cause defects such as blue spots,broken grids,missing corners,and wear.This defect will reduce power generation efficiency and reduce the service life.Therefore,it is of great significance to study the surface defects of solar cells and conduct regular inspections.The main work is as follows:Firstly,this paper takes photovoltaic panels with surface defects as the research object,and uses image processing technology to realize the detection of defects on the surface of solar cells.As there are dense parallel grid lines on the surface of solar cells,which are difficult to detect,which will seriously affect the accuracy of detection.This paper proposes a method combining Hough transform and Canny edge detection to remove grid interference.According to the two cases of complete removal of grid line and incomplete removal of grid line,image processing was carried out respectively.The grid line was initially positioned and deleted by Hough transformation.When the grid lines are completely removed,the operations of edge breakpoint connection,internal filling and smoothing are performed in sequence to obtain the target defect map;when the grid lines are not completely removed,the gray-scale equalization map is used for low-pass filtering and morphological expansion.It can be completely deleted.After test statistics,the average error of the improved detection method in this paper can reach 1.90%.Secondly,Experiments were carried out on monolithic photovoltaics,3×1photovoltaic arrays and 3×2 photovoltaic arrays,respectively,to study the effect of defect ratio on photovoltaic output.The study found that the open circuit voltage of the three groups of experiments will slowly decrease as the proportionof defects gradually increases.The change trend of short-circuit current is also similar.At first,the proportion of defects decreases slowly,but the proportion of short-circuit current will drop sharply within a certain range.Finally,the proportion of defects tends to be flat as the proportion of defects increases.Among them,when a single photovoltaic panel has surface defects,the proportion of current abrupt defects is 3%?4%;when a photovoltaic panel with a series relationship has surface defects,the proportion of current sudden defects will increase exponentially(the multiple is Series number);when there are surface defects in the photovoltaic panels with a parallel relationship,the proportion of defects with sharp current changes will also double(the multiple is the total number of photovoltaic panels).In addition,the fitting model of the defect ratio and the open circuit voltage is a fourth-order polynomial fitting.The better fitting model of the defect ratio and the short circuit current is the Boltzmann function,and the fitting accuracy can reach more than 0.99.Finally,according to the Matlab GUI,this paper designed a computer user interface of solar cell surface defect detection to display the picture,and it greatly improved the above methods of operability,intuitive and universal.