This paper aims to improve the generalization capability of feature extraction scheme by introducing a micro-cracks detection method based on self-learning features. Hence, such a design is ideally suited for an automated production environment and rapid in-line inspection process. This simplifies the detection process as complicated algorithms are avoided. Compared with photoluminescence imaging, the proposed system produces much cleaner images since it is less susceptible to artifacts in solar cell such as grain boundaries, surface structure, dislocations, cosmetic scratches, etc. The presence of microcrack along the transmission path of the transflected light causes an abrupt change in the illumination intensity, thereby creating a shadow representation of the defect. As a result, an area in the vicinity of the incident light is illuminated from beneath the surface. In this case, an incident light enters a solar cell and is diffusely scattered in multiple directions due to the crystalline material and surface roughness of a solar cell. Solving this problem leads to a design of an imaging system based on light transflection principle. State-of-the-art system that uses the photoluminescence technique for microcrack detection suffers from one major drawback-this method is sensitive to microcrack as well as noise. One major problem encountered during the manufacture of crystalline silicon solar cells is microcrack.
0 Comments
Leave a Reply. |
AuthorWrite something about yourself. No need to be fancy, just an overview. ArchivesCategories |