Research On Position Detection And Recognition Method Of Photovoltaic Cells Based On Machine Vision

Nowadays,the energy problem in the world is becoming more and more hot issues in the world.The traditional energy reserves are gradually reduced,and the environmental pollution is becoming increasingly serious.All countries in the world are eager to find new energy to replace the traditional energy.Solar energy,with its abundant reserves and no pollution to the environment,makes countries all over the world deeply study and develop solar energy resources.With the rapid development of photovoltaic industry,as a supporting facility for the development and utilization of photovoltaic industry,the research and development of photovoltaic cells are also concerned.In order to ensure the quality of the photovoltaic cells,it is a very important step to test the integrity of the cell in the photovoltaic cells automation production line.In the detection of whether the photovoltaic cells are completed or damaged,if the photovoltaic cells are in the process of loading,it is necessary to ensure that the placement position of the photovoltaic cells will not deflect.Once the position is deflected,it needs to be corrected immediately,so as to keep the normal welding process.At the same time,if the photovoltaic cells before welding are damaged,the system will identify them as unqualified raw materials,and the system will sort them.In this paper,machine vision technology is mainly used to detect and correct the position of photovoltaic cells,and to identify and sort the damaged cells,which lays a good foundation for the normal welding process of the following cells.This paper mainly studies the following parts:(1)In order to complete the position detection and correction of photovoltaic cells,a set of recognition scheme with machine vision system as the key point was designed and built in the laboratory environment.Around the optical components in the battery information recognition equipment,because there is a certain attitude deviation in the specific installation operation,the camera was calibrated by using the black and white chessboard method,and the camera was analyzed after the relationship between the machine coordinate system and the world coordinate system,the distortion parameters of the camera are studied and discussed.(2)In order to solve the computer's storage and calculation requirements for the photovoltaic cells,the matching background image is used to carry out effective segmentation and processing,and the method of eliminating the interference caused by unrelated images to the research is eliminated as far as possible;the mathematical morphology method is used to remove the fine grid lines of the photovoltaic cell images,which can accurately extract the photovoltaic cell images while solving the problems of large storage and calculation edge information.In order to improve the accuracy and speed of position detection of photovoltaic cells,Hough transform is adopted.Aiming at the corresponding characteristic straight line of the processed image recognition,multiple weighted mean method is used to obtain the central coordinates and deflection angles of the photovoltaic cells,and affine transform method is used to correct the position of the photovoltaic cells.Set up detailed coordinates and measurement error information,demonstrate detailed coordinates and specific error data of deflection angle measurement,and carry out effective comparison and research on the obtained experimental parameters.(3)In order to detect the location information and defect information of the damaged photovoltaic cells,the image of the template cells are obtained by moving the center position coordinate of the photovoltaic cell images.By comparing with the relevant parameters of the template photovoltaic cell images,we can judge whether there is any defect on the surface of the photovoltaic cells to be tested.