A clustering based approach to perceptual image hashing

Vishal Monga, A. Banerjee, B. L. Evans

Research output: Contribution to journalArticle

92 Citations (Scopus)

Abstract

A perceptual image hash function maps an image to a short binary string based on an image's appearance to the human eye. Perceptual image hashing is useful in image databases, watermarking, and authentication. In this paper, we decouple image hashing into feature extraction (intermediate hash) followed by data clustering (final hash). For any perceptually significant feature extractor, we propose a polynomial-time heuristic clustering algorithm that automatically determines the final hash length needed to satisfy a specified distortion. We prove that the decision version of our clustering problem is NP complete. Based on the proposed algorithm, we develop two variations to facilitate perceptual robustness versus fragility tradeoffs. We validate the perceptual significance of our hash by testing under Stirmark attacks. Finally, we develop randomized clustering algorithms for the purposes of secure image hashing.

Original languageEnglish (US)
Pages (from-to)68-79
Number of pages12
JournalIEEE Transactions on Information Forensics and Security
Volume1
Issue number1
DOIs
StatePublished - Mar 1 2006

Fingerprint

Clustering algorithms
Hash functions
Watermarking
Heuristic algorithms
Authentication
Feature extraction
Computational complexity
Polynomials
Testing

All Science Journal Classification (ASJC) codes

  • Safety, Risk, Reliability and Quality
  • Computer Networks and Communications

Cite this

@article{c9dab2c081f2443bbd55181c0a17eae6,
title = "A clustering based approach to perceptual image hashing",
abstract = "A perceptual image hash function maps an image to a short binary string based on an image's appearance to the human eye. Perceptual image hashing is useful in image databases, watermarking, and authentication. In this paper, we decouple image hashing into feature extraction (intermediate hash) followed by data clustering (final hash). For any perceptually significant feature extractor, we propose a polynomial-time heuristic clustering algorithm that automatically determines the final hash length needed to satisfy a specified distortion. We prove that the decision version of our clustering problem is NP complete. Based on the proposed algorithm, we develop two variations to facilitate perceptual robustness versus fragility tradeoffs. We validate the perceptual significance of our hash by testing under Stirmark attacks. Finally, we develop randomized clustering algorithms for the purposes of secure image hashing.",
author = "Vishal Monga and A. Banerjee and Evans, {B. L.}",
year = "2006",
month = "3",
day = "1",
doi = "10.1109/TIFS.2005.863502",
language = "English (US)",
volume = "1",
pages = "68--79",
journal = "IEEE Transactions on Information Forensics and Security",
issn = "1556-6013",
publisher = "Institute of Electrical and Electronics Engineers Inc.",
number = "1",

}

A clustering based approach to perceptual image hashing. / Monga, Vishal; Banerjee, A.; Evans, B. L.

In: IEEE Transactions on Information Forensics and Security, Vol. 1, No. 1, 01.03.2006, p. 68-79.

Research output: Contribution to journalArticle

TY - JOUR

T1 - A clustering based approach to perceptual image hashing

AU - Monga, Vishal

AU - Banerjee, A.

AU - Evans, B. L.

PY - 2006/3/1

Y1 - 2006/3/1

N2 - A perceptual image hash function maps an image to a short binary string based on an image's appearance to the human eye. Perceptual image hashing is useful in image databases, watermarking, and authentication. In this paper, we decouple image hashing into feature extraction (intermediate hash) followed by data clustering (final hash). For any perceptually significant feature extractor, we propose a polynomial-time heuristic clustering algorithm that automatically determines the final hash length needed to satisfy a specified distortion. We prove that the decision version of our clustering problem is NP complete. Based on the proposed algorithm, we develop two variations to facilitate perceptual robustness versus fragility tradeoffs. We validate the perceptual significance of our hash by testing under Stirmark attacks. Finally, we develop randomized clustering algorithms for the purposes of secure image hashing.

AB - A perceptual image hash function maps an image to a short binary string based on an image's appearance to the human eye. Perceptual image hashing is useful in image databases, watermarking, and authentication. In this paper, we decouple image hashing into feature extraction (intermediate hash) followed by data clustering (final hash). For any perceptually significant feature extractor, we propose a polynomial-time heuristic clustering algorithm that automatically determines the final hash length needed to satisfy a specified distortion. We prove that the decision version of our clustering problem is NP complete. Based on the proposed algorithm, we develop two variations to facilitate perceptual robustness versus fragility tradeoffs. We validate the perceptual significance of our hash by testing under Stirmark attacks. Finally, we develop randomized clustering algorithms for the purposes of secure image hashing.

UR - http://www.scopus.com/inward/record.url?scp=33645004781&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=33645004781&partnerID=8YFLogxK

U2 - 10.1109/TIFS.2005.863502

DO - 10.1109/TIFS.2005.863502

M3 - Article

AN - SCOPUS:33645004781

VL - 1

SP - 68

EP - 79

JO - IEEE Transactions on Information Forensics and Security

JF - IEEE Transactions on Information Forensics and Security

SN - 1556-6013

IS - 1

ER -