Benner
محمد رضا عبد الزهرة الكعبي ( مدرس مساعد )
كلية
[email protected]
 
 
 
Hybrid Correlative Measures for High-Performance Face Recognition with Intensity Adjustment
تحميل
بحث النوع:
علوم التخصص العام:
Mohammed R.A.M. Hammoodi اسم الناشر:
اسماء المساعدين:
Jour of Adv Research in Dynamical & Control Systems, Vol. 10, No. 10, 2018 الجهة الناشرة:
Jour of Adv Research in Dynamical & Control Systems, Vol. 10, No. 10, 2018  
2018 سنة النشر:

الخلاصة

The proposed method presents two efficient measures for face recognition. The approach is based on a proper weighting of (correlation coefficients) between corresponding face regions. The regions are a right eye, left eye, nose, and mouth as defined by Viola-Jones Algorithm. In addition, the measures have incorporated the holistic face region with a maximal weight. The recognition performance measure used is the distance between the reference image and the next confusing image in the database, where a better measure gives a bigger distance, hence better confidence in the recognition process. The proposed measures have been compared with structural similarity measure (SSIM), where superior performance has been obtained. Illumination adjustment of face images has been considered via differential thresholding that may enhance recognition. The recognition process assumed one face image as a reference image to be recognized in a database that includes the same reference person but with a different face pose (image).