•  
  •  
 

Turkish Journal of Electrical Engineering and Computer Sciences

DOI

10.3906/elk-2002-6

Abstract

Image hashing is one of the most celebrated techniques regarding the discipline of image forensics, imageretrieval, image indexing, content verification, and zero watermarking. For such sensitive and complex problems,generation of a unique and robust image hash is an utmost prerequisite for an image identifier driven from the perceptualcontents of an image. As a design perspective, it is essential for an image hash to have robustness and optimizeddiscriminative capability. We propose a robust image hashing technique by acquiring perceptual features based on anovel distance magnitude profile utilizing color pixel incongruity among the contiguous pixels, as well as producing astructural image for discrimination. The combination of both results provide a state-of-the-art robust profile that hase?icient discriminative capacity and is impervious to usual signal processing violations.

First Page

648

Last Page

662

Share

COinS