Issues in Information Hiding Transform Techniques

Abstract

This paper gives a review of prevailing transform embedding techniques for information hiding, and discusses the robustness and detectability of two specific methods. Spatial embedding inserts messages into image pixels. Transform embedding embeds a message by modifying selected frequency coefficients of the cover image. Ideally, the message embedded with transform techniques is more robust to different image processing attacks and more transparent to human visual systems than that of spatial domain techniques. Transform embedding techniques that we are particularly interested in are the Discrete Fourier Transform (DFT), the Discrete Cosine Transform (DCT), and the Discrete Wavelet Transform (DWT). Detection of the hidden message and robust embedding are two of the principal objectives of our information hiding research. We show that a watermarked image that is perceptually invisible in the spatial domain may fail our detectability test. Our approach to detectability is based on DFT domain analysis Embedding a message under the least significant bits (LSB) of pixels (LSB embedding) is commonly used. We investigate ways to improve the robustness of LSB embedding and propose a method for protecting embedded data against two specific forms of filtering.

Open PDF

Document Details

Document Type
Technical Report
Publication Date
May 20, 2002
Accession Number
ADA401948

Entities

People

  • Liwu Chang

Organizations

  • United States Naval Research Laboratory

Tags

Communities of Interest

  • Energy and Power Technologies

DTIC Thesaurus Topics

  • Coding
  • Coefficients
  • Detection
  • Discrete Fourier Transforms
  • Dynamic Range
  • Filters
  • Filtration
  • Fourier Analysis
  • Fourier Series
  • Frequency
  • Frequency Bands
  • Frequency Domain
  • Frequency Response
  • Image Compression
  • Image Processing
  • Military Research
  • Wavelet Transforms

Fields of Study

  • Computer science
  • Engineering

Readers

  • Cybersecurity.
  • Forest Ecology
  • Image Processing and Computer Vision.