A study of a multiscaled morphing interpolation technique for texture synthesis and image watermarking

博士 === 國立中興大學 === 資訊科學與工程學系所 === 98 === The morphing interpolation techniques are commonly used tools in computer graphics and image processing. They have been applied to many applications, such as view morphing, texture design, and morph animation. In this dissertation, we apply the morphing interp...

Full description

Bibliographic Details
Main Authors: Chao-Hung Lai, 賴昭宏
Other Authors: 吳俊霖
Format: Others
Language:en_US
Published: 2010
Online Access:http://ndltd.ncl.edu.tw/handle/81412666721323388256
Description
Summary:博士 === 國立中興大學 === 資訊科學與工程學系所 === 98 === The morphing interpolation techniques are commonly used tools in computer graphics and image processing. They have been applied to many applications, such as view morphing, texture design, and morph animation. In this dissertation, we apply the morphing interpolation techniques to temporal texture synthesis and image watermarking against local random geometric attacks. In the application of temporal texture synthesis, the proposed algorithm is further extended to render ocean surface. Temporal texture can be used to describe a wide variety of natural phenomena such as fire, smoke, and water. In this dissertation, we present a novel algorithm for temporal texture synthesis. This algorithm is simple and it requires only a static texture image as input to produce a plausible sequence with inexhaustible and quasi-periodic properties. We first introduce the basis sequence generation procedure. All the frames in the synthesized basis sequence are toroidal. Then the probability matrix and explicit links are employed to generate an inexhaustible sequence with quasi-periodic property. The quality of the synthesized temporal textures can be improved by the morphing interpolation techniques. In addition, the proposed method allows users to simply control the motion parameters and interactively render the composite landscape animations. The proposed morphing technique is also able to solve the problem of morphing between chaotic textures with obscure features. Several examples including scenes for cloud, fire, and water are presented to demonstrate that the proposed algorithm is simple, efficient, and controllable for synthesize temporal textures. Next, we extend the application for temporal texture synthesis and present another novel algorithm for synthesizing ocean surface with an illumination model. The proposed method requires two static images as input, one for reproducing the detailed ocean surface and the other for reconstructing the main structure waveform. The preprocessing stage synthesizes the toroidal textures and reconstructs the height fields for detailed water waves and the main structure waveform. The rendering stage uses the trivial texture mapping method to perform fast rendering, or uses a partial ray-marching method to achieve more realistic results without significantly increasing computational costs. The proposed method also simulates the specular effect to render more realistic ocean scenes, which is a great way to produce ocean landscape using only the photo textures. Experimental results demonstrate that the proposed method can easily and effectively produce plausible ocean scenes. Finally, we apply the multiscaled morphing technique to the application for image watermarking against the local geometric attack which is recently considered as the public enemy of most watermarking schemes. The watermark is embedded and retrieved using an amplitude modulation watermarking scheme. This work gives consideration to all the fidelity, robustness, and capacity. Moreover, the watermark bits act as implicit anchors to help correct local geometric distortions. In the proposed registration algorithm, an automatically multi-scaled neighborhood-matching approach is used to quickly and efficiently estimate local displacement, and a multi-resolution framework is superimposed to estimate both small and large local geometric distortions. Experimental results demonstrate the robustness and efficiency of the proposed algorithm.