Benner
أ.هند رستم شعبان ( أستاذ )
كلية علوم الحاسوب والرياضيات - الحاسوب
[email protected]
07810986626
 
 
 
, “An Efficient Information-Theoretic Measure for Image Similarity,” First International Conference on Scientific Studies and Humanities, University College of Humanities (UCH), Najaf, Iraq, 19-20 April 2016.
بحث النوع:
علوم التخصص العام:
Farah M. M. Altufaili اسم الناشر:
Zahir M. Hussain, Hind Rustum Mohammed اسماء المساعدين:
First International Conference on Scientific Studies and Humanities, الجهة الناشرة:
University College of Humanities (UCH), Najaf, Iraq, 19-20 April 2016  
2016 سنة النشر:

الخلاصة

ABSTRACT A new image information theoretic image similarity measure is proposed for grey scale image. The proposed measure is based on equation that used SSIM and HSSIM that base on joint histogram. The new measure that called SITS outperforms existing measures in detecting image similarity at low PSNR under Gaussian noise. The new measure SITS can detect similarity with confidence at least 20% increase than existing measures that used joint histogram alone, especially at low PSNR. Keywords: Information Theoretic (IT), Joint Histogram, Structural Similarity (SSSIM), Image Processing.