Improved Block Truncation Coding Using Optimized Error Diffusion and Speed-Oriented Halftone Watermarking Using Implicit Angle Lookup Strategy
| Autor: | Shih-hung Chou, 周士閎 |
|---|---|
| Rok vydání: | 2013 |
| Druh dokumentu: | 學位論文 ; thesis |
| Popis: | 101 In this thesis, two contributions are delivered, including improved block truncation coding using optimized error diffusion and speed-oriented halftone watermarking using implicit angle lookup strategy. The first proposed method is an improved halftoning-based Block Truncation Coding (BTC) using optimized error diffusion. First, the error-diffused block truncation coding (EDBTC) scheme adopts the two extreme values, maximum and minimum, within a block to form the quantization levels. Yet, this induces the unpleasant visual impulse noises. Thus, an adaptive quantization levels selection strategy is proposed to ease this problem, where the adjustable parameters are added to further modify the quantization levels, and the optimized parameters for different blocks are investigated. In addition, the error kernel used in the EDBTC is, in fact, for the typical halftoning, rather than BTC-based. Consequently, the optimization on error kernel for BTC-based scheme is proposed, in which the Genetic Algorithm (GA) is adopted. Experimental results demonstrate that the proposed method can achieve the highest image quality with the same compression ratio former BTC methods. The second proposed method is a speed-oriented halftone watermarking using implicit angle lookup strategy. Herein, the computational complexity can be significantly reduced. In encoder, the Direct Binary Search (DBS) is employed to generate the reference table to ensure the output is in halftone format. Subsequently, a number of optimized compressive tables with various texture angles are established for subsequent table lookup. In decoder, the Least-Mean-Square (LMS) enlarges the differences among those phenotypes of the embedded angles and the required number of dimensions for each angle. Finally, the naive Bayes classifier is employed to collect the probability information for classifying various angles. As documented in the experimental results, good image quality and correct detect rate can be obtained simultaneously. Moreover, a high processing efficiency of 0.6 milliseconds for an image of size 512x512 can also be achieved, which can thus increase the commercial competitive strength in printing market, in particular the security issue is well addressed. |
| Databáze: | Networked Digital Library of Theses & Dissertations |
| Externí odkaz: |
|