Digital Watermarking and Steganography

Digital Watermarking and Steganography

Kuanchin Chen (Western Michigan University, USA)
DOI: 10.4018/978-1-60566-014-1.ch054
OnDemand PDF Download:


Sharing, disseminating, and presenting data in digital format is not just a fad, but it is becoming part of our life. Without careful planning, digitized resources could easily be misused, especially those that are shared across the Internet. Examples of such misuse include use without the owner’s permission, and modification of a digitized resource to fake ownership. One way to prevent such behaviors is to employ some form of copyright protection technique, such as digital watermarks. Digital watermarks refer to the data embedded into a digital source (e.g., images, text, audio, or video recording). They are similar to watermarks in printed materials as a message inserted into the host media typically becomes an integral part of the media. Apart from traditional watermarks in printed forms, digital watermarks may also be invisible, may be in the forms other than graphics, and may be digitally removed.
Chapter Preview

Information Hiding, Steganography, And Watermarking

To many people, information hiding, steganography, and watermarking refer to the same set of techniques to hide some form of data. This is true in part because these terms are closely related to each other, and sometimes they are used interchangeably.

Information hiding is a general term that involves message embedding in some host media (Cox, Miller, & Bloom, 2002). The purpose of information hiding is to make the information imperceptible or to keep the existence of the information secret. Steganography means “covered writing,” a term derived from the Greek literature. Its purpose is to conceal the very existence of a message. Digital watermarking, however, embeds information into the host documents, but the embedded information may be visible (e.g., a company logo), or invisible (in which case, it is similar to steganography).

Steganography and digital watermarking differ in several ways. First, the watermarked messages are related to the host documents (Cox et al., 2002). An example is the ownership information inserted into an image. Second, digital watermarks do not always have to be hidden. See Taylor, Foster, and Pelly (2003) for the applications of visible watermarks. However, visible watermarks are typically not considered steganography by definition (Johnson & Jajodia, 1998). Third, watermarking requires additional “robustness” in its algorithms. Robustness refers to the ability of a watermarking algorithm to resist from removal or manipulation attempts (Acken, 1998; Craver, Perrig, & Petitcolas, 2000). This characteristic deters attackers by forcing them to spend an unreasonable amount of computation time and/or by inflicting an unreasonable amount of damage to the watermarked documents in the attempts of watermark extraction. Figure 1 shows that there are considerable overlaps in the meaning and even the application of the three terms. Many of the algorithms in use today are, in fact, shared among information hiding, steganography, and digital watermarking. The difference relies largely on “the intent of use” (Johnson & Jajodia, 1998). Therefore, discussions in the rest of the article on watermarking also apply to steganography and information hiding, unless specifically mentioned otherwise.

Figure 1.

Information hiding, steganography, and digital watermarking

To be consistent with the existing literature, a few common terms are described below. Cover work refers to the host document (text, image, multimedia, or other media content) that will be used to embed data. The data to be embedded is called the watermark and may be in the form of text, graphic, or other digital format. The result of this embedding is called a stego-object.


Characteristics Of Effective Watermarking Algorithms

Watermarking algorithms are not created equal. Some will not survive from simple image processing operations, while others are robust enough to deter attackers from some forms of modifications. Effective and robust image watermarking algorithms should meet the following requirements:

Key Terms in this Chapter

Least Significant Bit (LSB): An LSB refers to the last or the right-most bit in a binary number. It is called LSB because changing its value will not dramatically affect the resulting number.

Stego-Object: This is the cover work with a watermark inserted or embedded.

Cover Work: The host media in which a message is to be inserted or embedded

Information Hiding: The umbrella term referring to techniques of hiding various forms of messages into cover work

Steganography: Covered writing; it is a study of concealing the existence of a message.

Cryptography: A study of making a message secure through encryption; secret key and public key are the two major camps of cryptographic algorithms. In secret key cryptography, one key is used for both encryption and decryption, while in public key cryptography, two keys (public and private) are used.

Digital Watermarking: This concerns the act of inserting a message into a cover work. The resulting stego-object can be visible or invisible.

Complete Chapter List

Search this Book: