1、 44 7 2010 M7上 海 交 通 大 学 学 报JOU RNAL OF SH ANGH AI JIAOTONG UNIVERSITYVol. 44 No. 7 Jul. 2010 l : 2009- 09-08Te: Ar( 1985- ) , 3, , V 3,1V Y;_;“dZ . ( “ ) , 3, q, V 3 =,( T el. ) : 021-54745803; E- mail: x u-lin sjtu. edu. cn.cI|: 1006-2467( 2010) 07-0926-05M atlabm) / MYA J柳效辉, 徐 林, 肖晨江, 曹建明, 肖 娇(
2、ZYv “ ? , Z200240)K 1: 针对硅太阳电池缺陷的检测问题, 利用非接触红外检测方法,通过 CCD 相机获取硅太阳电池的电致发光图像,利用Matlab软件对缺陷部位图像首先进行预处理,然后进行图像分割a特征提取和图像识别,最后判断检测的硅太阳电池的缺陷.结果表明,该方法能够有效识别硅太阳电池的缺陷.1oM: 硅太阳电池; 缺陷; 图像处理; 识别ms |: TP 391. 41 DS : AIdentifying the Crack of SiliconSolar CellsBased onMatlab Image ProcessingLIUXiao-hui, XULin, X
3、IAOChen-jiang, CAOJian-ming, XIAOJiao( Solar Energy Institute, Department of Physics, Shanghai Jiaotong U niversity, Shanghai 200240, China)Abstract: Aiming at the m easure problem of crack of silicon solar cells, an infrared non-contact detectionmethod w as used. The electrolum inescence image of s
4、ilicon solar cells w as captured by CCD camera. Imagepre- processing, image segm entation, featur e extraction and image identification w ere analyzed based onMatlab image processing. T he r esult proves that, using the pr inciple, cr ack of silicon solar cells can be de-tected.Key words: silicon so
5、lar cells; crack; im age processing; identification M , “; 334)Z1 (i) = 2* Z1 (i)elseZ1 (i) = 0. 5* Z1( i)endendZ= waverec2( Z1 , s,-sym4. ) %s“ figure, imshow( Z, )title(-9 0 % m count = count+ 1;endendendS= count %d9 m 9 s, 7 :H ATB H,5 V U, p99 .N H, H, :2; H, :1,V UL = Ne+ 2N0 (2)T:NeaN0sYH (8Z_
6、) 0)P2 = jif ( ( P2 - P1 ) = = 1)% 0)P2 = iif ( ( P2 - P1 ) = = 1)% Z_ Nx = Nx + 1endP1 = P2endendendNs = sum( sum( BW) ) %9 9 Nd = Ns - Nx - Ny %9 L= sqrt( 2) * Nd + Nx + Ny %9 3. 2 MY m+(B %) , L. , 9 T 8C= L2/(4PS) (3)T:C;L;S . A 9 , VS = 554,L= 192. 953,5C=5.367.C3 7 V 9 ,79 C= 5.4, V .4 Matlab qA m) , iT.N$ ,s m+4 |ZEm MYZE,K V _ JBH .h . 3m) # M .:S , 2006. 8 Anil K. Jain. 3m) $ M .: bv, 2006. 9 . 3m) / j _“d D .?:?v/#1, 2003.930 上 海 交 通 大 学 学 报44
