【作者】 孙东明；

【导师】 申铉京；

【作者基本信息】 吉林大学 ， 软件工程， 2013， 硕士

【摘要】 太阳能电池阵已经是现代科技发展不可或缺的一部分，太阳能电池片是太阳能电池阵的重要组成部分，如果电池片破损将直接影响到电池阵的输出功率。因此，对太阳能电池片的检测是非常重要的一个环节。现如今，大量采用的检测方法是目测法，检测人员使用专业的检测显微镜对太阳能电池片进行目测，这种方法不仅劳动量大，而且检测人员的视力、主观意识、身体状况、疲劳过度等客观因素都对检测结果都会有一定的影响。所以研发卫星太阳电池片裂纹检测装置是非常必要且迫切的，具有重要的经济和社会价值。如今，图像检测技术正在高速发展，科学技术水平也在日益提高，检测的方法也在渐渐由传统人工调查的方式向高科技的检测方式过渡。本论文的论述以此为出发点，对太阳能电池片的裂纹检测进行了进一步研究。本文使用的是一种自动扫描检测的方法，设计并实现了太阳能电池片裂纹检测系统中的图像处理模块。该模块方法首先针对图像采集模块提供的每张太阳能电池片图像进行二值化处理，其目的是为了消除光影的影响，并显现出栅极线特征；其次，根据二值化后的太阳能电池片图像具有特征，对其进行垂直边界的查找；第三，通过搜索栅极线的横截线段计算出栅极线平均间隔，再根据拍摄的图像特征，可获取栅极线首列位置；第四，确定了首列位置再加上栅极线间隔，就可对栅极线进行填充，并且可以精确切割左右边界；第五，使用更加严格的阈值进行二次处理，使裂纹特征显现出来；最后，去除三角形电极，通过噪声与裂纹的差别来区别噪声和裂痕信息，查找出裂纹之后，计算出可能被栅极线断开的裂纹并连接。本论文首先在第一章介绍了研究背景以及意义，并且相关研究概况进行了简介；第二章介绍了检测系统的硬件组成以及裂纹检测系统的各个组成模块；在第三章中详细叙述了系统图像处理模块的具体算法流程，这也是裂纹检测系统设计的重点和难点；第四章首先介绍了系统的操作界面，并对图像处理模块进行了测试实验，对实验结果进行了分析。实验证明，图像处理模块的设计和实现，能够查找出图像中裂纹的位置，满足了系统的要求。因此完成的太阳能裂纹检测系统可以很好代替人工检测，应用到生产工作中去。本文对于裂纹检测的研究具有一定的价值。更多还原

【Abstract】 The solar array is already an integral part of the development of modern scienceand technology, the solar cells are important parts of the solar arrays.The cells whichare damaged will directly affect the power output of the solar array. Therefore,detection is a very important aspect of the solar cell sheet. Today, a large number ofdetect methods use visual methods, testing personnel conducting visual detectionmicroscope solar cells, these methods are not only a large amount of labors andvisions, but also subjective awareness of the testing personnel, physical condition,fatigue, and other objective factors will have a certain influence on the test results.Therefore, research and development of satellite solar cell chip crack detection deviceis very necessary and urgent, also be good for the economic and society. Today, theimage detection technology is developing rapidly in scientific and technological level,also is increasing detection methods are gradually transition to the high-tech detectionby traditional manual survey.The paper discusses this as a starting point, and further study of the crackdetection of solar cells. This article uses an automatic scan detection method anddesigns and implements solar cells crack detection system image processing module.Each solar cell slice image of the module method for image acquisition moduleprovides first binarization processing, and its purpose is to eliminate the influence oflight and shadow, and show a gate line characteristic. Secondly, according to thebinarized solar cell sheet image having the characteristics of its vertical boundarylookup. Third, by searching the segment of a cross section of the gate lines iscalculated the average interval of the gate lines, according to the shooting of theimage feature, to obtain the first column position of the gate line. Fourth determinethe first column position of the gate line coupled with the upper gate line interval, canbe filled, and can be precisely cut the left and right borders. Fifth, the use of morestringent the threshold value for secondary processing, so that the crack characteristicsmanifested. Last, to remove triangles electrode noise to distinguish the difference between noise and cracks and cracks, find out the cracks, calculate the rift may bedisconnected from the gate line and connect. In the first chapter, the paper introducesthe research background and significance, and research profile. The second chapterdescribes the various components of the detection system composed of hardware andcrack detection system module; In the third chapter,system is described in detail inthe specific algorithm flow of the image processing module, which is the crackdetection system design emphasis and difficulty; Chapter four first introduces theoperation of the system interface and an image processing module test experiment, theexperimental results were analyzed. Experimental results show that the design andimplementation of image processing module, it is able to find out the location of thecrack in the image to meet the requirements of the system. Therefore, the solar crackdetection system well in place of the manual inspection, and production work. Thispaper has a certain value for crack detection.更多还原