ادغام ويژگي هاي رنگ و بافت به منظور تشخيص خويشاوندي از روي تصاوير چهره

Color and texture feature fusion for facial kinship verification

ﺳﯿﺴﺘﻢ ﺗﺸﺨﯿﺺ ﺧﻮﯾﺸﺎوﻧﺪي ﺑﺎ ﺗﺠﺰﯾﻪ و ﺗﺤﻠﯿﻞ ﺗﺼﺎوﯾﺮ ﭼﻬﺮه دو ﻓﺮد، ﺧﻮﯾﺸﺎوﻧﺪي ﯾﺎ ﻋـﺪم ﺧﻮﯾﺸـﺎوﻧﺪي آن دو را ﻣﺸـﺨﺺ ﻣﯽ ﮐﻨـﺪ. ﺑﺮاي ﺗﺸﺨﯿﺺ راﺑﻄﻪ ﺧﻮﯾﺸﺎوﻧﺪي وﯾﮋﮔﯽ ﻫﺎي ﻣﺨﺘﻠﻒ را ﻣﯽ ﺗﻮان از ﺗﺼـﻮﯾﺮ ﭼﻬـﺮه دو ﻓـﺮد اﺳـﺘﺨﺮاج ﻧﻤـﻮد. در اﯾـﻦ ﻣﻘﺎﻟـﻪ ﺑـﺎ ﺑﺮرﺳـﯽ وﯾﮋﮔﯽ ﻫﺎي ﺑﺎﻓﺖ، رﻧﮓ و ادﻏﺎم ﻣﻮﺛﺮ اﯾﻦ وﯾﮋﮔﯽ ﻫﺎ و ﻫﻤﭽﻨﯿﻦ ﺑﺮرﺳﯽ ﭼﻨﺪ ﻃﺒﻘﻪ ﺑﻨﺪي ﮐﻨﻨﺪه ﻣﺨﺘﻠﻒ، ﯾﮏ ﺳﯿﺴﺘﻢ ﮐﺎرا ﺑﺮاي ﺗﺸـﺨﯿﺺ رواﺑﻂ ﺧﻮﯾﺸﺎوﻧﺪي ﻧﺴﻞ اول ) ﭘﺪر - ﭘﺴـﺮ، ﭘـﺪر - دﺧﺘـﺮ، ﻣـﺎدر - ﭘﺴـﺮ و ﻣـﺎدر - دﺧﺘـﺮ( اراﺋـﻪ ﺷﺪه اﺳـﺖ. در اﯾـﻦ راﺳـﺘﺎ دو روﯾﮑـﺮد ﭘﯿﺸﻨﻬﺎدي ﺑﺮرﺳﯽ ﺷﺪه اﺳـﺖ: )1( ادﻏـﺎم وﯾﮋﮔﯽ ﻫـﺎي ﻣـﺆﺛﺮ و ﺑﺮرﺳـﯽ ﻃﺒﻘﻪ ﺑﻨﺪي ﮐﻨﻨـﺪه ﻣﺨﺘﻠـﻒ ﺑـﺮاي ﺗﺸـﺨﯿﺺ ﺧﻮﯾﺸـﺎوﻧﺪي و )2( اﺳــﺘﻔﺎده از ﻣﺘﺮﯾــﮏ ﯾــﺎدﮔﯿﺮي NRML ﺑــﻪ ﻣﻨﻈﻮر ﺗﻮﻟﯿــﺪ ﺑــﺮدار وﯾﮋﮔــﯽ ﻣﺘﻤﺎﯾﺰ ﮐﻨﻨــﺪه ﺟﻬــﺖ اﻓــﺰاﯾﺶ ﮐــﺎراﯾﯽ ﺗﺸــﺨﯿﺺ ﺧﻮﯾﺸــﺎوﻧﺪي. روش ﻫﺎي ﭘﯿﺸﻨﻬﺎدي ﺑﺮاي دو ﭘﺎﯾﮕﺎه داده KinFaceW-I و KinFaceW-II در ﺣﺎﻟﺖ ﻫﺎي ﻣﺨﺘﻠﻒ ﺗﺤﻠﯿـﻞ و ارزﯾـﺎﺑﯽ ﺷـﺪه اﻧﺪ. ﻧﺘـﺎﯾﺞ ارزﯾﺎﺑﯽﻫﺎ ﻧﺸﺎن ﻣﯽدﻫﺪ، ادﻏﺎم وﯾﮋﮔﯽ ﻫﺎ و اﺳﺘﻔﺎده از ﻣﺘﺮﯾﮏ NRML ﺑﻪ ﺧﻮﺑﯽ ﺗﻮاﻧﺴﺘﻪ اﺳﺖ ﻋﻤﻠﮑﺮد ﺳﯿﺴﺘﻢ ﺗﺸﺨﯿﺺ ﺧﻮﯾﺸـﺎوﻧﺪي را ﺑﻬﺒﻮد دﻫﺪ. ﻋﻼوه ﺑﺮ دو روﯾﮑﺮد ﭘﯿﺸﻨﻬﺎدي، اﺳﺘﺨﺮاج وﯾﮋﮔﯽ از ﮐﻞ ﺗﺼﻮﯾﺮ و ﻫﻤﭽﻨﯿﻦ ﺑﻪ ﺻﻮرت ﺑﻠﻮﮐﯽ از ﺗﺼـﻮﯾﺮ، ﺑﺮرﺳـﯽ ﺷـﺪه و ﻧﺘﺎﯾﺞ آن اراﺋﻪ ﺷﺪه اﺳﺖ. ﻧﺘﺎﯾﺞ ﺑﻪ دﺳﺖ آﻣﺪه ﺣﺎﮐﯽ از آن اﺳﺖ ﮐﻪ اﺳﺘﺨﺮاج وﯾﮋﮔﯽ ﺑﻪ ﺻﻮرت ﺑﻠﻮﮐﯽ ﻣﯽ ﺗﻮاﻧﺪ در ﺑﻬﺒﻮد ﻧﺘﯿﺠـﻪ ﻧﻬـﺎﯾﯽ ﺗﺸﺨﯿﺺ ﺧﻮﯾﺸﺎوﻧﺪي ﻣﻮﺛﺮ واﻗﻊ ﺷﻮد.

The kinship Verification system analyzes the facial features of two people to determine whether they are related or not. To identify the kinship, different features can be extracted from the faces. In this paper, to evaluate a kinship verification system for the first-generation kinship (father-son, father-daughter, mother-son, and mother-daughter), texture and color features are tested, and feature fusion, as well as examining several different classifiers is considered. In this regard, two proposed approaches have been proposed: (1) fusing effective features and evaluate different classifiers for kinship verification and (2) using NRML metric learning to generate a distinctive feature vector to increase kinship verification efficiency. The proposed methods for the two databases KinFaceW-I and KinFaceW-II have been analyzed and evaluated in different cases. The results of the evaluations show that the fusion of features and the use of NRML metric learning have been able to improve the performance of the kinship verification system. In addition to the two proposed approaches, feature extraction from the whole image as well as image blocks is proposed and the results are presented. The results indicate that using the block-wise method for feature extraction can be effective in improving the final kinship verification results.