通过对真彩图像与灰色MRI切片的亮度和纹理等特征的配准而进行灰色图像的彩色化,可以得到有效的而且符合真实人体器官组织的真彩仿真图像.着重研究将粒子群算法与点邻域亮度纹理分布和局部颜色结构分布相结合而应用于Welsh图像彩色化算法.通过粒子群寻优迭代方程的各参数的自适应调节和控制提升其性能,并将点邻域分布的颜色特征和彩色局部分布的颜色特征通过多种搜索算法如遗传算法和粒子群算法及其修正算法进行了适应度函数最优值搜索分析比较,其彩色化的MRI切片的三维重构模型能多层次清晰地反映器官组织的分布和构造.%Researches transferring a gray MRI slice into a colorful one that is valid and matches the human organizations by registering the features of brightness and textures of a true color image and a gray image.The Welsh color transferring algorithm is particularly probed based on the combination of particle swarm optimization and the feature of brightness within bit neighborhood and local color distribution.An improved particle swarm optimization (PSO) algorithm by self-adapting regulating and control of parameters of the optimal formula to promote the performance.A pixel's color neighbor distribution and local color structure distribution properties by exhaustion,random selection,genetic algorithm (GA) ,PSO and improved PSO and their integration for a fitness optimum searching and their analysis and comparisoa Reconstructed 3D color model from the colored slices could clearly reveal multilevel structure and distribution.
展开▼
