东南大学学报（自然科学版）2016，Vol.46Issue(5)：957-963,7.DOI:10.3969/j.issn.1001-0505.2016.05.010
基于字典学习的超分辨率显微CT图像重建
Super-resolution image reconstruction for micro-CT based on dictionary learning
摘要
Abstract
To improve the spatial resolution of reconstructed images for micro computed tomography (micro-CT),a super-resolution image reconstruction algorithm based on dictionary learning is pro-posed.First,the reconstructed image grid is refined and the area weight model is used to achieve ac-curate modeling of the projection process.Then,high quality images are selected as the training samples.The K-means singular value decomposition (K-SVD)algorithm is adopted to train the im-age dictionary.On the basis of the image dictionary,the orthogonal matching pursuit algorithm is used to implement sparse representation of the reconstructed image,which is introduced into the ob-jective function of the reconstruction algorithm as a sparse constraint.Finally,the gradient descent method is adopted to solve the objective function.The experimental results show that compared to the conventional interpolation-based super-resolution reconstruction algorithms,the proposed algo-rithm has advantages on the image contrast and the edge preservation,and retains more high frequen-cy information of images,effectively improving the spatial resolution of the reconstructed images.关键词
超分辨率重建/字典学习/面积权值/微计算机断层扫描技术Key words
super-resolution reconstruction/dictionary learning/area weight/micro computed tomography(micro-CT)分类
信息技术与安全科学引用本文复制引用
姚佳丽,李中源,吴华珍,李光,罗守华..基于字典学习的超分辨率显微CT图像重建[J].东南大学学报（自然科学版）,2016,46(5):957-963,7.基金项目
国家自然科学基金资助项目（61127002，61179035）． （）
