Dissertation > Industrial Technology > Automation technology,computer technology > Computing technology,computer technology > Computer applications > Information processing (information processing) > Pattern Recognition and devices > Image recognition device

The Research on Texture Synthesis Technology from Cloud Theory & Been Evolution Genetic Algorithm

Author SunTao
Tutor XuWeiHong
School Changsha University of Science and Technology
Course Applied Computer Technology
Keywords Texture synthesis Texture Feature Cloud model Bee Evolution Genetic Algorithm Best Optimal Sample Block
CLC TP391.41
Type Master's thesis
Year 2011
Downloads 10
Quotes 0
Download Dissertation

Texture synthesis is a common technology used in creating complex scenes and graphing realistic images. Among all the methods of texture synthesis, sample-based texture synthesis has been developed in recent years as a new texture synthesis technology. In order to synthesize any size and high-quality images of similar texture, this method uses the small pieces of texture image as the input data. During the process of texture synthesis, meeting the real-time need in terms of synthesis speed and forming a fully automatic synthesis process are regarded as the research targets.Based on the study of sample-based texture synthesis technology,the thesis presents improvements of the existing theory and algorithms. The main points are as follows:1.In order to solve the problem of the edge seams more efficiently during the texture synthesis, the thesis proposes a new texture algorithm based on cloud theory, which first makes the statistical descriptors of the texture fuzzy, and then uses algebraic inverse operation to generate the 2D vector cloud of texture feature. Through calculating the cloud distance among different images, the algorithm realizes the 3D cloud growth of the texture features and finally completes the texture synthesis with several repeated operations. Experimental results show that this algorithm is better than the classical Chaotic Particle Swarm Optimization algorithm in terms of high accuracy and fast synthetic speed.2.Based on Bee Evolutionary Genetic Algorithm and Wang Tiles Texture Synthesis Algorithm, the thesis proposes an improved method of the best optimal sample block selection. It shrinks the searching scope and therefore raises the speed. Besides, the algorithm keeps the texture details more complete for small pieces texture synthesis. Simulation experiment results verify the feasibility of this algorithm.

Related Dissertations
More Dissertations