×ÔÊÊÓ¦Â˲¨Æ÷Éè¼Æ¼°MatlabʵÏÖ£¬¸½³ÌÐò´úÂë

άÄÉ×ÔÊÊÓ¦Â˲¨Æ÷Éè¼Æ¼°MatlabʵÏÖ

Õª Òª

±¾ÎÄ´ÓËæ»úÔëÉùµÄÌØÐÔ³ö·¢£¬·ÖÎöÁË´«Í³Â˲¨ºÍ×ÔÊÊÓ¦Â˲¨»ù±¾¹¤×÷Ô­ÀíºÍÐÔÄÜ£¬ÒÔ¼°Â˲¨¼¼ÊõµÄÏÖ×´ºÍ·¢Õ¹Ç°¾°¡£È»ºóϵͳ²ûÊöÁË»ù±¾Î¬ÄÉÂ˲¨Ô­ÀíºÍ×ÔÊÊÓ¦Â˲¨Æ÷µÄ»ù±¾½á¹¹Ä£ÐÍ£¬½Ó×ÅÔÚ´Ë»ù´¡ÉϽáºÏ×Ͻµ·¨Òý³öLMSËã·¨¡£ÔÚMSE×¼ÔòÏ£¬Éè¼ÆÁËÒ»¸ö¶¨³¤µÄ×ÔÊÊÓ¦×îС¾ù·½ºáÏòÂ˲¨Æ÷£¬²¢Í¨¹ýMATLAB±à³ÌʵÏÖ¡£½Ó×ÅÓÃͼÏñ¸´Ô­À´ÑéÖ¤¸ÃÂ˲¨Æ÷µÄÐÔÄÜ£¬½á¹û±íÃ÷ͼÏñµÄÖÊÁ¿ÔÚMSE×¼Ôòϵõ½ÁËÃ÷ÏԵĸÄÉÆ¡£×îºó·ÖÎö±È½ÏÁË×ÔÊÊÓ¦LMSÂ˲¨ºÍƵÓòάÄɵݹéÂ˲¨Ö®¼äµÄÐÔÄÜ¡£±¾ÎÄ»¹¶ÔMATLABÀïÃæµÄ×ÔÊÊӦάÄÉÂ˲¨º¯Êýwiener2½øÐÐÁ˼òµ¥·ÖÎö¡£

¹Ø¼ü×Ö£ºÍË»¯Í¼Ïñ άÄÉÂ˲¨ ×ÔÊÊÓ¦Â˲¨

×Ͻµ·¨ LMS

Abstract

This paper analyses the basic work theory, performance of traditional filter and adaptive filter based on the property of random noise, and introduce the status quo and the foreground of filter technology. Then we explain basic theory of wiener filter and basic structure model of adaptive filter, and combine the method of steepest descent to deduce the LMS. Afterward according to the MSE rule, we design a limited length transversal filter, and implement by MATLAB. And then we validate performance of adaptive LMS filter by restoring images, Test result show that the quality of the degrade images were improved under the rule of MSE. Finally, we compare the performance of adaptive LMS filter and iterative wiener filter.

We also simply analyses the wiener2 () which is a adaptive filter in MATLAB.

Keywords: degrade image£»wiener filter£»adaptive filter£»ADF£»LMS algorithm

Ŀ¼

1 Ð÷ÂÛ¡­¡­¡­¡­¡­¡­¡­¡­¡­¡­¡­¡­¡­¡­¡­¡­¡­¡­¡­¡­¡­¡­¡­¡­¡­¡­¡­¡­¡­¡­¡­1 1. 1 ÒýÑÔ¡­¡­¡­¡­¡­¡­¡­¡­¡­¡­¡­¡­¡­¡­¡­¡­¡­¡­¡­¡­¡­¡­¡­¡­¡­¡­¡­¡­1 1. 2 Ñо¿Ä¿±ê¼°ÏÖ×´¡­¡­¡­¡­¡­¡­¡­¡­¡­¡­¡­¡­¡­¡­¡­¡­¡­¡­¡­¡­¡­¡­¡­1

1. 2 .1 ͼÏñ¸´Ô­¼¼ÊõµÄÄ¿±ê¡­¡­¡­¡­¡­¡­¡­¡­¡­¡­¡­¡­¡­¡­¡­¡­¡­1 1. 2 .2 ͼÏñ¸´Ô­¼¼ÊõµÄÑо¿ÏÖ×´¡­¡­¡­¡­¡­¡­¡­¡­¡­¡­¡­¡­¡­¡­¡­1

2 ÀíÂÛ»ù´¡ ¡­¡­¡­¡­¡­¡­¡­¡­¡­¡­¡­¡­¡­¡­¡­¡­¡­¡­¡­¡­¡­¡­¡­¡­¡­¡­¡­¡­3 2. 1 »ù±¾×ÔÊÊÓ¦Â˲¨Æ÷µÄÄ£¿é½á¹¹¡­¡­¡­¡­¡­¡­¡­¡­¡­¡­¡­¡­¡­¡­¡­¡­¡­3 2. 2 »ù±¾Î¬ÄÉÂ˲¨Ô­Àí¡­¡­¡­¡­¡­¡­¡­¡­¡­¡­¡­¡­¡­¡­¡­¡­¡­¡­¡­¡­¡­¡­4 3 ×ÔÊÊÓ¦Â˲¨Ô­Àí¼°Ëã·¨ ¡­¡­¡­¡­¡­¡­¡­¡­¡­¡­¡­¡­¡­¡­¡­¡­¡­¡­¡­¡­¡­6 3.1 ºáÏòÂ˲¨½á¹¹µÄ×ϽµËã·¨¡­¡­¡­¡­¡­¡­¡­¡­¡­¡­¡­¡­¡­¡­¡­¡­¡­7

3.1.1 ×ϽµËã·¨µÄÔ­Àí¡­¡­¡­¡­¡­¡­¡­¡­¡­¡­¡­¡­¡­¡­¡­¡­¡­7 3.1.2 ×ϽµËã·¨Îȶ¨ÐÔ¡­¡­¡­¡­¡­¡­¡­¡­¡­¡­¡­¡­¡­¡­¡­¡­¡­10 3.2 LMSÂ˲¨Ô­Àí¼°Ëã·¨¡­¡­¡­¡­¡­¡­¡­¡­¡­¡­¡­¡­¡­¡­¡­¡­¡­¡­¡­¡­11

3.2.1 ´Ó×ϽµËã·¨µ¼³öLMSËã·¨ ¡­¡­¡­¡­¡­¡­¡­¡­¡­¡­¡­¡­11 3.2.2 »ù±¾LMSËã·¨µÄʵÏÖ²½Öè ¡­¡­¡­¡­¡­¡­¡­¡­¡­¡­¡­¡­¡­¡­11 3.2.3 »ù±¾LMSËã·¨µÄʵÏÖÁ÷³Ìͼ ¡­¡­¡­¡­¡­¡­¡­¡­¡­¡­¡­¡­¡­12 3.2.4 LMSËã·¨µÄMatlabʵÏÖ ¡­¡­¡­¡­¡­¡­¡­¡­¡­¡­¡­¡­¡­¡­¡­12 3.2.5 wiener2()µÄÔ­Àí ¡­¡­¡­¡­¡­¡­¡­¡­¡­¡­¡­¡­¡­¡­¡­¡­¡­12 3.2.6 LMSÐÔÄÜ·ÖÎö¡ª¡ª×ÔÊÊÓ¦ÊÕÁ²ÐÔ¡­¡­¡­¡­¡­¡­¡­¡­¡­¡­¡­13

4 Matlab ʵÑé½á¹û ¡­¡­¡­¡­¡­¡­¡­¡­¡­¡­¡­¡­¡­¡­¡­¡­¡­¡­¡­¡­¡­¡­14 4.1.LMSÂ˲¨Æ÷µÄÊÕÁ²ÐÔ ¡­¡­¡­¡­¡­¡­¡­¡­¡­¡­¡­¡­¡­¡­¡­¡­¡­¡­14 4.2.LMSÂ˲¨Æ÷ºÍƵÓòµü´úάÄÉÂ˲¨Æ÷µÄÐÔÄÜ±È½Ï ¡­¡­¡­¡­¡­¡­¡­¡­16 5 ×ܽᡭ¡­¡­¡­¡­¡­¡­¡­¡­¡­¡­¡­¡­¡­¡­¡­¡­¡­¡­¡­¡­¡­¡­¡­¡­¡­¡­¡­¡­¡­18 ÖÂл ¡­¡­¡­¡­¡­¡­¡­¡­¡­¡­¡­¡­¡­¡­¡­¡­¡­¡­¡­¡­¡­¡­¡­¡­¡­¡­¡­¡­¡­¡­¡­19 ²Î¿¼ÎÄÏ× ¡­¡­¡­¡­¡­¡­¡­¡­¡­¡­¡­¡­¡­¡­¡­¡­¡­¡­¡­¡­¡­¡­¡­¡­¡­¡­¡­¡­¡­20 ¸½Â¼A ¡­¡­¡­¡­¡­¡­¡­¡­¡­¡­¡­¡­¡­¡­¡­¡­¡­¡­¡­¡­¡­¡­¡­¡­¡­¡­¡­¡­¡­¡­21 ¸½Â¼B ¡­¡­¡­¡­¡­¡­¡­¡­¡­¡­¡­¡­¡­¡­¡­¡­¡­¡­¡­¡­¡­¡­¡­¡­¡­¡­¡­¡­¡­¡­22 ¸½Â¼C ¡­¡­¡­¡­¡­¡­¡­¡­¡­¡­¡­¡­¡­¡­¡­¡­¡­¡­¡­¡­¡­¡­¡­¡­¡­¡­¡­¡­¡­¡­27

1 Ð÷ÂÛ

1.1ÒýÑÔ

ÈËÀà´«µÝÐÅÏ¢µÄÖ÷Ҫý½éÊÇÓïÑÔºÍͼÏñ¡£¾Ýͳ¼Æ£¬ÔÚÈËÀà½ÓÊܵÄÐÅÏ¢ÖУ¬Ìý¾õÐÅÏ¢Õ¼20£¥£¬ÊÓ¾õÐÅÏ¢Õ¼60£¥£¬ÆäËüÈçζ¾õ¡¢´¥¾õ¡¢Ðá¾õ×ܵļÓÆðÀ´²»¹ýÕ¼20£¥£¬ËùÒÔͼÏñÐÅÏ¢ÊÇÊ®·ÖÖØÒªµÄÐÅÏ¢[1]¡£È»¶ø£¬ÔÚͼÏñµÄ»ñÈ¡ºÍͼÏñÐźŵĴ«Êä¹ý³ÌÖУ¬Í¼ÏñÐźÅÖв»¿É±ÜÃâµÄ»ìÈë¸÷ÖÖ¸÷ÑùµÄËæ»úÔëÉù£¬Ôì³ÉͼÏñʧÕ棨ͼÏñÍË»¯£©¡£Ôì³ÉÈËÀàËù»ñÈ¡µÄÐÅÏ¢ºÍʵ¼ÊÊÇÓÐÆ«²îµÄ£¬³ÉΪÈËÀà´ÓÍâ½ç»ñȡ׼ȷÐÅÏ¢µÄÕÏ°­¡£Òò´Ë£¬¶ÔͼÏñÐźÅÖеÄËæ»úÔëÉùµÄÒÖÖÆ´¦ÀíÊÇͼÏñ´¦ÀíÖзdz£ÖØÒªµÄÒ»Ï×÷¡£

ÔÚͼÏñµÄ»ñÈ¡ºÍ´«Êä¹ý³ÌÖÐËù»ìÈëµÄÔëÉù£¬Ö÷ÒªÀ´Ô´ÓÚͨÐÅϵͳÖеĸ÷ÖÖ¸÷ÑùµÄÔëÉù£¬¸ù¾ÝͨÐÅÔ­Àí¼°Í³¼Æ·½ÃæµÄ֪ʶ£¬¿ÉÒÔÖªµÀÔÚͨÐÅϵͳÖÐËùÓöµ½µÄÐźźÍÔëÉù£¬´ó¶àÊý¾ù¿ÉÊÓΪƽÎȵÄËæ»ú¹ý³Ì[15]¡£ÓÖÓС°¸ß˹¹ý³ÌÓÖ³ÆÕý̬Ëæ»ú¹ý³Ì£¬ËüÊÇÒ»ÖÖÆÕ±é´æÔÚºÍÖØÒªµÄËæ»ú¹ý³Ì£¬ÔÚͨÐÅÐŵÀÖеÄÔëÉù£¬Í¨³£ÊÇÒ»ÖÖ¸ß˹¹ý³Ì£¬¹ÊÓֳƸß˹ÔëÉù¡£Òò´Ë£¬ÔÚ´ó¶àÊýµÄÇé¿öÏ£¬ÎÒÃÇ¿ÉÒÔ°ÑÔì³ÉͼÏñʧÕæµÄÔëÉù¿ÉÊÓΪ¹ãÒåƽÎȸß˹¹ý³Ì¡£

±¾ÎÄÕë¶ÔͼÏñÐźÅÖлìÈëµÄËæ»úÔëÉù£¬ÔÚÔõÑù°ÑÏÖÓеÄÂ˲¨Ëã·¨Ó¦Óõ½Êµ¼ÊµÄͼÏñ¸´Ô­ÖÐÈ¥µÄÎÊÌâÉÏÌá³öÁ˽â¾ö·½·¨£¬²¢ÇÒÓ¦ÓÃMatlab Èí¼þ±à³Ì¶ÔͼÏñ½øÐд¦Àí¡£

1.2Ñо¿Ä¿±ê¼°ÏÖ×´

1.2.1ͼÏñ¸´Ô­¼¼ÊõµÄÄ¿±ê

ΪÁË´Óº¬ÓÐÔëÉùµÄÊý¾ÝÖÐÌáÈ¡ÎÒÃÇËù¸ÐÐËȤµÄ¡¢½Ó½ü¹æ¶¨ÖÊÁ¿µÄͼÏñ£¬ÎÒÃÇÐèÒªÉè¼ÆÒ»¸öϵͳÂú×㣺µ±ÐźÅÓëÔëÉùͬʱÊäÈëʱ£¬ÔÚÊä³ö¶ËÄܽ«Ðźž¡¿ÉÄܾ«È·µØÖØÏÖ³öÀ´£¬¶øÔëÉùÈ´Êܵ½×î´óÒÖÖÆ£¬¼´×î¼ÑÂ˲¨Æ÷¡£ 1.2.2ͼÏñ¸´Ô­¼¼ÊõµÄÑо¿ÏÖ×´

Ä¿Ç°µÄͼÏñ¸´Ô­¼¼Êõ£¬¼´È¥ÔëµÄÂ˲¨¼¼Êõ¿ÉÒÔ·ÖΪÁ½´óÀࣺ´«Í³Â˲¨ºÍÏÖ´úÂ˲¨¡£´«Í³Â˲¨¼¼ÊõÊǽ¨Á¢ÔÚÒÑÖªÓÐÓÃÐźź͸ÉÈÅÔëÉùµÄͳ¼ÆÌØÐÔ£¨×ÔÏà¹Øº¯Êý»ò¹¦ÂÊÆ×£©µÄ»ù´¡ÉϵÄÔëÉùÈ¥³ý£»ÏÖ´úÂ˲¨¼¼ÊõÔòÊDz»ÐèÒªÖªµÀͼÏñµÄÏÈÑé֪ʶ£¬Ö»ÊǸù¾Ý¹Û²âÊý¾Ý£¬¼´¿É¶ÔÔëÉù½øÐÐÓÐЧÂ˳ý¡£

ÔçÔÚ20ÊÀ¼Í40Äê´ú£¬¾Í¶ÔƽÎÈËæ»úÐźŽ¨Á¢ÁËάÄÉÂ˲¨ÀíÂÛ¡£¸ù¾ÝÓÐÓÃÐÅ

ºÅºÍ¸ÉÈÅÔëÉùµÄͳ¼ÆÌØÐÔ£¨×ÔÏà¹Øº¯Êý»ò¹¦ÂÊÆ×£©£¬ÒÔÏßÐÔ×îС¾ù·½Îó²î£¨MSE£©¹À¼Æ×¼ÔòËùÉè¼ÆµÄ×î¼ÑÂ˲¨Æ÷£¬³ÆΪάÄÉÂ˲¨Æ÷¡£ÕâÖÖÂ˲¨Æ÷ÄÜ×î´ó³Ì¶ÈµÄÂ˳ý¸ÉÈÅÔëÉù£¬ÌáÈ¡ÓÐÓÃÐźš£µ«ÊÇ£¬µ±ÊäÈëÐźŵÄͳ¼ÆÌØÐÔÆ«ÀëÉè¼ÆÌõ¼þ£¬ÔòËü¾Í²»ÔÙÊÇ×î¼ÑµÄÁË£¬ÕâÔÚʵ¼ÊÓ¦ÓÃÖÐÊܵ½ÁËÏÞÖÆ¡£µ½60Äê´ú³õ£¬ÓÉÓڿռ似ÊõµÄ·¢Õ¹£¬³öÏÖÁË¿¨¶ûÂüÂ˲¨ÀíÂÛ£¬¼´ÀûÓÃ״̬±äÁ¿Ä£ÐͶԷÇƽÎÈ¡¢¶àÊäÈë¶àÊä³öËæ»úÐòÁÐ×÷×îÓŹÀ¼Æ¡£¿¨¶ûÂüÂ˲¨Æ÷¼È¿ÉÒÔ¶ÔƽÎȵĺÍƽÎȵÄËæ»úÐźÅ×÷ÏßÐÔ×î¼ÑÂ˲¨£¬Ò²¿ÉÒÔ×÷Ϊ·ÇÏßÐÔÂ˲¨[2]¡£

È»¶øÖ»ÓÐÔÚ¶ÔÐźźÍÔëÉùµÄͳ¼ÆÌØÐÔÒÑÖªµÄÇé¿öÏ£¬ÕâÁ½ÖÖÂ˲¨Æ÷²ÅÄÜ»ñµÃ×îÓŽ⡣ÔÚʵ¼ÊµÄÓ¦ÓÃÖУ¬ÍùÍùÎÞ·¨µÃµ½ÕâЩͳ¼ÆÌØÐÔµÄÏÈÑé֪ʶ£¬»òÕßͳ¼ÆÌØÐÔÊÇËæʱ¼ä±ä»¯µÄ£¬Òò´Ë£¬ÕâÁ½ÖÖÂ˲¨Æ÷¾ÍʵÏÖ²»ÁËÕæÕýµÄ×î¼ÑÂ˲¨¡£

Widrow B.ºÍHoffÓÚ1967ÄêÌá³öµÄ×ÔÊÊÓ¦Â˲¨ÀíÂÛ£¬¿ÉʹÔÚÉè¼Æ×ÔÊÊÓ¦Â˲¨Æ÷ʱ²»ÐèÒªÊÂÏÈÖªµÀ¹ØÓÚÊäÈëÐźźÍÔëÉùµÄͳ¼ÆÌØÐÔµÄ֪ʶ£¬ËüÄܹ»ÔÚ×Ô¼ºµÄ¹¤×÷¹ý³ÌÖÐÖð½¥¹À¼Æ³öËùÐèµÄͳ¼ÆÌØÐÔ£¬²¢ÒÔ´ËΪÒÀ¾Ý×Ô¶¯µ÷Õû×Ô¼ºµÄ²ÎÊý£¬ÒÔ´ïµ½×î¼ÑÂ˲¨Ð§¹û¡£Ò»µ©ÊäÈëÐźŵÄͳ¼ÆÌØÐÔ·¢Éú±ä»¯£¬ËüÓÖÄܹ»¸ú×ÙÕâÖֱ仯£¬×Ô¶¯µ÷Õû²ÎÊý£¬Ê¹Â˲¨Æ÷ÐÔÄÜÖØдﵽ×î¼Ñ¡£

×ÔÊÊÓ¦Â˲¨Æ÷×Ô¶¯µ÷½Ú²ÎÊý¿ÉÒÔͨ¹ý¸÷ÖÖ²»Í¬µÄµÝÍÆËã·¨À´ÊµÏÖ£¬ÓÉÓÚËü²ÉÓõÄÊDZƽüµÄËã·¨£¬Ê¹µÃʵ¼Ê¹À¼ÆÖµºÍÀíÂÛÖµÖ®¼ä±ØÈ»´æÔÚ²î¾à£¬Ò²¾ÍÔì³ÉÁË×ÔÊÊÓ¦Â˲¨ÎÊÌâûÓÐΨһµÄ½â¡£ÒÀÕÕ¸÷ÖÖµÝÍÆËã·¨µÄÌص㣬ÎÒÃÇ°ÑËüÓ¦ÓÃÓÚ²»Í¬µÄ³¡ºÏ¡£ÏÖÔÚ¹ãΪӦÓõÄ×ÔÊÊÓ¦Â˲¨·½·¨Ö÷ÒªÊÇ»ùÓÚÒÔϼ¸ÖÖ»ù±¾ÀíÂÛ£¬ÔÙÈںϵÝÍÆËã·¨µ¼³öÀ´µÄ£º

£¨1£© »ùÓÚάÄÉÂ˲¨ÀíÂ۵ķ½·¨

άÄÉÂ˲¨ÊÇÔÚ×îС¾ù·½Îó²î×¼ÔòÏÂͨ¹ýÇó½âάÄÉ¡ª»ô·ò·½³ÌÀ´½â¾öÏßÐÔ×îÓÅÂ˲¨ÎÊÌâµÄ¡£»ùÓÚάÄÉÂ˲¨Ô­Àí£¬ÎÒÃÇÀûÓÃÏà¹ØµÄ˲ʱֵͨ¹ýÔÚ¹¤×÷¹ý³ÌÖеÄÖð²½µ÷Õû²ÎÊý±Æ½üÐźŵÄͳ¼ÆÌØÐÔ£¬ÊµÏÖ×îÓÅÂ˲¨¡£ÓÉ´Ë£¬ÎÒÃǵõ½Ò»ÖÖ×î³£ÓõÄËã·¨¡ª¡ª×îС¾ù·½Ëã·¨£¬¼ò³ÆLMSËã·¨¡£

£¨2£© »ùÓÚ¿¨¶ûÂüÂ˲¨ÀíÂ۵ķ½·¨

¿¨¶ûÂüÂ˲¨ÊÇÏßÐÔÎÞÆ«×îС·½²îÂ˲¨µÝÍÆÂ˲¨£¬ËüÄÜʹÂ˲¨Æ÷¹¤×÷ÔÚƽÎȵĻò·ÇƽÎȵĻ·¾³£¬µÃµ½×îÓŽ⡣ÀûÓÿ¨¶ûÂüÂ˲¨ÀíÂ۵ĵÝÍÆÇó½â·¨µ¼³ö×ÔÊÊÓ¦Â˲¨Æ÷¸üÐÂȨʸÁ¿µÃ²»Í¬µÝÍÆËã·¨¡£±ÈLMSËã·¨Óм«¿ìµÄÊÕÁ²ËÙÂÊ£¬¿ÉÊǼÆË㸴