Joint Gaussian Conditional Random Field for Multi-Focus Image Fusion

• Main Authors: Sin-yi Jiang, 江欣怡
• Other Authors: Kai-lung Hua
• Format: Others
• Language: zh-TW
• Published: 2014
• Online Access: http://ndltd.ncl.edu.tw/handle/49185806745861400671

碩士 === 國立臺灣科技大學 === 資訊工程系 === 102 === Multi-focus image fusion aims to combine a set of images that are captured from the same scene but with different focuses for producing another sharper image. The critical issue in the design of multi-focus image fusion algorithms is to evaluate the local content information of the input images. Joint Gaussian conditional random field (JGCRF) for multi-focus image fusion is proposed in this paper. First, the features and naive weight in each input image are defined as random variables, and we desire to obtain the weight corresponding to each pixel. Then, the dependency relationship between random variables which follows the joint Gaussian distribution is connected with edges. The graph is an undirected graph that consist of edges and nodes (variables), the relationship of the undirected graph is represented through the JGCRF model. As the optimal solution is obtained by applying maximum a posteriori (MAP), the weight map of each focus images is obtained. Finally, multi-focus images are combined to a fused image containing completed and clear depth of field. Experimental results on several fusion of multi-focus image show that the proposed method can give good results.