ÄãºÃ£¬»¶Ó­À´µ½¾­¹ÜÖ®¼Ò [µÇ¼] [×¢²á]

ÉèΪÊ×Ò³ | ¾­¹ÜÖ®¼ÒÊ×Ò³ | Êղر¾Õ¾

»ùÓÚÈËÁ³Ê¶±ðµÄͼÏñÔ¤´¦ÀíËã·¨Ñо¿ÓëʵÏÖ_ͨÐŹ¤³ÌרҵÂÛÎÄ

·¢²¼Ê±¼ä£º2015-01-24 À´Ô´£ºÈË´ó¾­¼ÃÂÛ̳
ͨÐŹ¤³ÌרҵÂÛÎÄ Ä¿ ¼ Òý ÑÔ1 µÚ1Õ Ð÷ÂÛ2 1.1 ¿ÎÌâ±³¾°2 1.2 ÈËÁ³Ê¶±ð¼¼ÊõµÄÑо¿ÒâÒå2 1.2.1 ¸»ÓÐÌôÕ½ÐԵĿÎÌâ2 1.2.2 Ã沿¹Ø¼üÌØÕ÷¶¨Î»¼°ÈËÁ³2DÐÎ×´¼ì²â¼¼Êõ2 1.2.3 Ã沿¸Ð֪ϵͳµÄÖØÒªÄÚÈÝ3 1.3 ÈËÁ³Ê¶±ðµÄ¹úÄÚÍâ·¢Õ¹¸Å¿ö3 1.3.1 ¹úÍâ·¢Õ¹¸Å¿ö4 1.3.2 ¹úÄڵķ¢Õ¹¸Å¿ö4 µÚ2Õ ÈËÁ³Í¼ÏñÔ¤´¦Àí5 2.1 ÒýÑÔ5 2.2 ÈËÁ³Í¼Ïñ¿â5 2.3 ÈËÁ³µÄÔ¤´¦ÀíËã·¨6 2.3.1 ͼÏñ¹éÒ»»¯6 2.3.2 ͼÏñ¶þÖµ»¯9 2.4 ¹âÕÕ²¹³¥9 µÚ3Õ ÈËÁ³Í¼ÏñµÄÌØÕ÷ÌáÈ¡11 3.1 K-L±ä»»11 3.2 ÆæÒìÖµ·Ö½â(SVD)13 3.3 Ö÷ÌØÕ÷ÌáÈ¡µÄ»ù±¾Ô­Àí14 3.4 ÈËÁ³Í¼ÏñµÄÌØÕ÷ÌáÈ¡15 3.4.1 »ùÓÚPCAµÄÈËÁ³ÌØÕ÷ÌáÈ¡Ô­Àí16 3.4.2 ÈËÁ³Í¼ÏñÌØÕ÷ÌáÈ¡Ëã·¨µÄʵÏÖ18 3.5 ÑÛ¾¦¶¨Î»¹ý³Ì19 3.5.1 ²ÊÉ«¿Õ¼äÄ£Ð͵ÄÑ¡È¡20 3.5.2 ¹âÏß²¹³¥21 3.5.3 »ùÓÚÑÛ¾¦ÁÁ¶ÈÖµµÄÅòÕÍ22 3.5.4 ÑÛ¾¦µÄ¶¨Î»22 3.6 ×ìºÍ±Ç×Ó¶¨Î»23 3.6.1 Ã沿¼¸ºÎ¹ØϵµÄÓ¦ÓÃ24 3.6.2 ±Ç×Ó¶¨Î»24 3.6.3 ×춨λ25 µÚ4Õ ÈËÁ³Í¼ÏñÔ¤´¦ÀíµÄÀíÂÛʵÏÖ27 4.1 ͼÏñµÄÔ¤´¦Àí³õ²½ÊµÏÖ27 4.2 Ã沿ÌØÕ÷µÄ¶¨Î»µÄÀíÂÛʵÏÖ28 ½áÂÛÓëÕ¹Íû29 Ö л30 ²Î¿¼ÎÄÏ×31 ¸½Â¼A£ºÖ÷Òª³ÌÐò32 ¸½Â¼B£ºÒ»ÆªÒýÓõÄÍâÎÄÎÄÏ×¼°ÒëÎÄ34 ¸½Â¼C£ºÖØÒª²Î¿¼ÎÄÏ×Ìâ¼40 ÕªÒª ÈËÁ³Ê¶±ðÒòÆäÔÚ°²È«Ñé֤ϵͳ¡¢ÐÅÓÿ¨ÑéÖ¤¡¢Ò½Ñ§¡¢µµ°¸¹ÜÀí¡¢ÊÓƵ»áÒé¡¢ÈË»ú½»»¥¡¢ÏµÍ³¹«°²(×ﷸʶ±ðµÈ)µÈ·½ÃæµÄ¾Þ´óÓ¦ÓÃÇ°¾°¶øÔ½À´Ô½³ÉΪµ±Ç°Ä£Ê½Ê¶±ðºÍÈ˹¤ÖÇÄÜÁìÓòµÄÒ»¸öÑо¿Èȵ㡣 ÈËÁ³Ê¶±ðµÄ¹Ø¼ü¾ÍÊÇÈËÁ³Í¼ÏñÔ¤´¦Àí£¬Ô¤´¦ÀíЧ¹ûµÄºÃ»µÖ±½Ó¹Øϵ×ÅÈËÁ³Ê¶±ðµÄ½á¹û¡£±¾ÎÄÑо¿ÁËÒ»¸ö»ùÓÚPCAÈËÁ³Ê¶±ðµÄÈËÁ³Í¼ÏñÔ¤´¦Àí·½·¨£¬²ÉÓÃPCA·½·¨¾ÍÊÇÀûÓÃK-L±ä»»ºÍSVDµÃµ½Õý½»»ù£¬È»ºó¸ù¾ÝÖ÷ÌØÕ÷ÌáÈ¡µÄÔ­Àí£¬Ñ¡È¡½Ï´óÌØÕ÷Öµ¶ÔÓ¦µÄ»ùÏòÁ¿¡£Ä¿Ç°Ëã·¨½ö½öÕë¶Ôµ¥ÈËÕýÃæµÄͼÏñ£¬ÓкܴóµÄ¾ÖÏÞÐÔ¡£ ±¾ÎÄËù²ÉÓõķ½·¨£¬³ýÁ˶ÔËã·¨µÄÓÅ»¯Í⣬¸ü¼Ó×¢ÖØͼÏñÔ¤´¦ÀíµÄЧ¹û¡£ ¹Ø¼ü×Ö£ºÈËÁ³Ê¶±ð£»ÈËÁ³Ô¤´¦Àí£»¹âÕÕ²¹³¥£»K-L±ä»»£»Ã沿ÌØÕ÷¶¨Î» Research and Implementation Of Image Pre-processing Algorithm Based on Face Recognition Abstract Face recognition is important in surveillance and security, telecommunications, digital libraries , video meeting, and human-computer intelligent interactions. It has been a research focus of pattern recognition and artificial intelligence. The important of face recognition is face image preprocessing. The pretreatment directly effect the results of the face recognition. In this paper, we study and implement a face image preprocessing method based on PCA. PCA use K-L transform and SVD to get orthogonal basis, and then according to the principles of the main feature extraction to select the larger eigenvalues of the corresponding vector-based. The current algorithm is only for the single positive face image and it have limitation. The method in this paper uses algorithm optimization and the effects of image preprocessing to achieve the performance of the algorithm. Keyword£ºFace recognition£»Face pretreatment£»Light compensating£»K-L transform£»Characteristics of position
¾­¹ÜÖ®¼Ò¡°Ñ§µÀ»á¡±Ð¡³ÌÐò
  • ɨÂë¼ÓÈë¡°¿¼ÑÐѧϰ±Ê¼ÇȺ¡±
ÍƼöÔĶÁ
¾­¼ÃѧÏà¹ØÎÄÕÂ
±êÇ©ÔÆ
¾­¹ÜÖ®¼Ò¾«²ÊÎÄÕÂÍƼö