Texture Synthesis Algorithm Based On Quantum-behaved Particle Swarm Optimization

The goal of texture synthesis is to produce texture by human. It is an important research area in both computer graphics and image processing. Sample-based texture synthesis was a new technique proposed in recent years. It takes small texture image as input to synthesis same kind of texture but in arbitrary size. The goal of sample-based texture synthesis is high quality output image, synthesis speed that satisfies real time application and full automatic synthesis process. This technique is widely used in producing virtual reality scene and satisfies the great need in application areas such as movie industry and computer game. With the rapid development of computer 3D graphics ability and the great need of large size and high quality texture in recent years, sample-based texture synthesis became a hot research area.The main challenge of this area currently lies in synthesis speed and the quality of synthesized image. In order to solve these problems, this thesis aims at fast and high quality texture synthesis algorithm, several effective methods are also proposed. Main work and innovation points of this thesis include the following:Firstly, a new texture synthesis algorithm utilizing QPSO is firstly proposed in this thesis in order to achieve the global best of MRF on the basis of analyzing the advantages and disadvantages of this texture model. Experiments show that this algorithm can reduce the possibility of trapping into local minima which both accelerates the speed of texture synthesis process and improves the quality of output image.Secondly, bases on the traditional similarity measurement in texture synthesis, this thesis proposes a new texture synthesis algorithm using mutual information. Experiments prove that this new algorithm can improve the quality of synthesized image especially when there are noises or data-missing in the sample image.Thirdly, a new method of global feature recognition and extraction for structured texture which based on polygonal approximation of digital curves is firstly proposed in this thesis. This method can locate texture global characteristic by extracting the feature points of structured texture and half-random-half structured texture. This thesis also proposes a new patch-based texture synthesis utilizing characteristic locating.