On the design and analysis of quantization-based digital watermarking systems
In quantization-based digital watermarking, a watermark message is conveyed in the choice of different quantizers. Because of its advantage of rejecting interference from host signals, a quantization-based digital watermarking system has the property of high information embedding rate. In this thesi...
Lưu vào:
| Tác giả chính: | |
|---|---|
| Định dạng: | Luận án |
| Ngôn ngữ: | en_US |
| Thông tin xuất bản: |
University of Waterloo
2007
|
| Chủ đề: | |
| Truy cập trực tuyến: | http://ir.vnulib.edu.vn/handle/123456789/1198 |
| Từ khóa: |
Thêm từ khóa bạn đọc
Không có từ khóa, Hãy là người đầu tiên gắn từ khóa cho biểu ghi này!
|
| Tóm tắt: | In quantization-based digital watermarking, a watermark message is conveyed in the choice of different quantizers. Because of its advantage of rejecting interference from host signals, a quantization-based digital watermarking system has the property of high information embedding rate. In this thesis, we investigate the design, analysis and application of quantization-based digital watermarking systems.
As watermarked signals are likely to be transmitted/stored in compressed formats in most applications, we first look into the integration of quantization-based digital water-marking with compression-joint watermarking and compression (JWC), and demonstrate its advantage over separate watermarking and compression (SWC). In JWC, a key pro¬cess is quantization which embeds watermarks into a host signal while digitizing the host signal subject to the requirements on the embedding rate, compression rate, quantization distortion and robustness. Using fixed-rate quantization, optimum encoding and decoding schemes for JWC systems are designed and analyzed to maximize the robustness of the systems in the presence of additive Gaussian attacks under constraints on the compres¬sion rate and quantization distortion. Simulation results show that in comparison with SWC systems, optimum JWC systems indeed achieve significant performance gains in the distortion-to-noise ratio region of practical interest.
Inspired by the advantage of JWC over SWC, we then investigate the quantization-based digital watermarking in the application of image compression in order to expand the application scope of digital watermarking. We look into the design of a joint im¬age compression and blind watermarking system to maximize compression rate distortion performance while maintaining baseline JPEG decoder compatibility and satisfying the additional constraint imposed by watermarking. To achieve blind and robust watermark¬ing, a differential quantization watermarking (DQW) method is proposed. The embedded watermark can survive a class of standard JPEG recompression attacks. To maximize the compression rate distortion performance, an iterative algorithm is developed to jointly
optimize run-length coding, Huffman coding and quantization table selection subject to the additional constraint imposed by DQW. In addition to embedding a certain amount of watermark information within compressed bit streams, the proposed joint watermarking and compression algorithm is demonstrated to achieve efficient compression performance.
To further broaden the application scope of digital watermarking, we finally apply quantization-based digital watermarking to image quality assessment. In this research, we propose a novel reduced-reference (RR) image quality assessment method, in which extracted features of a reference image are embedded into the same image as hidden mes¬sages. When a distorted version of such an image is received, users can decode the hidden messages and use them to help evaluate the quality of the distorted image using an RR quality assessment method. To demonstrate the idea, a practical image quality assessment system is proposed, which employs a novel RR image quality assessment algorithm based on statistical models of natural images, and a quantization-based watermarking technique in the wavelet transform domain. Experimental results show that the proposed RR image quality assessment method can achieve the performance comparable to those of widely used full-reference image quality assessment methods which require access to original images as references. |
|---|