中文摘要：

针对典型相关分析没有充分利用样本的类标签信息且在相关子空间选择近邻时需人工设值的问题,提出一种改进的超分辨率重构方法.首先利用鉴别典型相关分析最大化高、低分辨率图像在投影空间下的相关性;其次在相关子空间重构时采用稀疏表示动态选择近邻样本,并逐步得到测试图像对应的高分辨率图像;最后加上残差图像得到最终的恢复图像.实验结果表明,该方法在视觉和峰值性噪比、结构相似性指标的测评值上都有更好的效果.

英文摘要：

Super-resolution image method based on canonical correlation analysis is a super-resolution reconstructionmethod in linear subspace,but canonical correlation analysis methods do not take full advantage of labelinformation of the training sample class.Therefore,this paper presents an improved image super-resolution reconstructionmethod,namely using canonical correlation analysis added label information to maximize correlationbetween high-resolution images and low-resolution images in projection space for taking full advantage ofclassification information of the training samples.When reconstructing in coherent subspace,the method ofsparse representation are used to choose the number of neighbors to increase the flexibility of the model,and then,converting the high resolution image corresponding to the low-resolution image test step by step.Experimentalresults show better results both visually and on the value of PSNR and SSIM.