Perceptual image hashing via feature points: Performance evaluation and tradeoffs

Vishal Monga, Brian L. Evans

Research output: Contribution to journalArticle

227 Citations (Scopus)

Abstract

We propose an image hashing paradigm using visually significant feature points. The feature points should be largely invariant under perceptually insignificant distortions. To satisfy this, we propose an iterative feature detector to extract significant geometry preserving feature points. We apply probabilistic quantization on the derived features to introduce randomness, which, in turn, reduces vulnerability to adversarial attacks. The proposed hash algorithm withstands standard benchmark (e.g., Stirmark) attacks, including compression, geometric distortions of scaling and small-angle rotation, and common signal-processing operations. Content changing (malicious) manipulations of image data are also accurately detected. Detailed statistical analysis in the form of receiver operating characteristic (ROC) curves is presented and reveals the success of the proposed scheme in achieving perceptual robustness while avoiding misclassification.

Original languageEnglish (US)
Pages (from-to)3452-3465
Number of pages14
JournalIEEE Transactions on Image Processing
Volume15
Issue number11
DOIs
StatePublished - Nov 1 2006

Fingerprint

Statistical methods
Signal processing
Detectors
Geometry

All Science Journal Classification (ASJC) codes

  • Software
  • Computer Graphics and Computer-Aided Design

Cite this

@article{cf7efa9386064a04bbdd815bc9928ab3,
title = "Perceptual image hashing via feature points: Performance evaluation and tradeoffs",
abstract = "We propose an image hashing paradigm using visually significant feature points. The feature points should be largely invariant under perceptually insignificant distortions. To satisfy this, we propose an iterative feature detector to extract significant geometry preserving feature points. We apply probabilistic quantization on the derived features to introduce randomness, which, in turn, reduces vulnerability to adversarial attacks. The proposed hash algorithm withstands standard benchmark (e.g., Stirmark) attacks, including compression, geometric distortions of scaling and small-angle rotation, and common signal-processing operations. Content changing (malicious) manipulations of image data are also accurately detected. Detailed statistical analysis in the form of receiver operating characteristic (ROC) curves is presented and reveals the success of the proposed scheme in achieving perceptual robustness while avoiding misclassification.",
author = "Vishal Monga and Evans, {Brian L.}",
year = "2006",
month = "11",
day = "1",
doi = "10.1109/TIP.2006.881948",
language = "English (US)",
volume = "15",
pages = "3452--3465",
journal = "IEEE Transactions on Image Processing",
issn = "1057-7149",
publisher = "Institute of Electrical and Electronics Engineers Inc.",
number = "11",

}

Perceptual image hashing via feature points : Performance evaluation and tradeoffs. / Monga, Vishal; Evans, Brian L.

In: IEEE Transactions on Image Processing, Vol. 15, No. 11, 01.11.2006, p. 3452-3465.

Research output: Contribution to journalArticle

TY - JOUR

T1 - Perceptual image hashing via feature points

T2 - Performance evaluation and tradeoffs

AU - Monga, Vishal

AU - Evans, Brian L.

PY - 2006/11/1

Y1 - 2006/11/1

N2 - We propose an image hashing paradigm using visually significant feature points. The feature points should be largely invariant under perceptually insignificant distortions. To satisfy this, we propose an iterative feature detector to extract significant geometry preserving feature points. We apply probabilistic quantization on the derived features to introduce randomness, which, in turn, reduces vulnerability to adversarial attacks. The proposed hash algorithm withstands standard benchmark (e.g., Stirmark) attacks, including compression, geometric distortions of scaling and small-angle rotation, and common signal-processing operations. Content changing (malicious) manipulations of image data are also accurately detected. Detailed statistical analysis in the form of receiver operating characteristic (ROC) curves is presented and reveals the success of the proposed scheme in achieving perceptual robustness while avoiding misclassification.

AB - We propose an image hashing paradigm using visually significant feature points. The feature points should be largely invariant under perceptually insignificant distortions. To satisfy this, we propose an iterative feature detector to extract significant geometry preserving feature points. We apply probabilistic quantization on the derived features to introduce randomness, which, in turn, reduces vulnerability to adversarial attacks. The proposed hash algorithm withstands standard benchmark (e.g., Stirmark) attacks, including compression, geometric distortions of scaling and small-angle rotation, and common signal-processing operations. Content changing (malicious) manipulations of image data are also accurately detected. Detailed statistical analysis in the form of receiver operating characteristic (ROC) curves is presented and reveals the success of the proposed scheme in achieving perceptual robustness while avoiding misclassification.

UR - http://www.scopus.com/inward/record.url?scp=33750344108&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=33750344108&partnerID=8YFLogxK

U2 - 10.1109/TIP.2006.881948

DO - 10.1109/TIP.2006.881948

M3 - Article

C2 - 17076404

AN - SCOPUS:33750344108

VL - 15

SP - 3452

EP - 3465

JO - IEEE Transactions on Image Processing

JF - IEEE Transactions on Image Processing

SN - 1057-7149

IS - 11

ER -