基于多字典学习和图像块映射超分辨率重建.docVIP

基于多字典学习和图像块映射超分辨率重建 　　摘要：针对单一冗余字典在稀疏表示图像超分辨率重建结果出现不清晰、伪影以及重建过程编码效率不高、运算时间过长的问题，提出一种基于多字典学习和图像块映射的超分辨率重建方法。该方法在传统稀疏表示的框架下，首先探索局部图像块的梯度结构信息，按梯度角度将训练样本块分类； 然后为每个子类样本集学习高低分辨率字典对，再结合最近邻思想应用生成的字典，为每个子类计算从低分辨率块到高分辨率块映射的函数；最后将重建过程简化为输入块和映射函数的乘积，在保证提高重建质量的同时减少了图像重建的时间。实验结果表明，所提算法在视觉效果有较大的提升，同时与锚点邻域回归算法相比，评价参数峰值信噪比（PSNR）平均提高约0.4dB。 　　关键词：稀疏表示；图像块分类；多字典学习；映射函数；邻域嵌入 　　中图分类号：TP391 文献标志码：A 　　Abstract：To overcome the disadvantages of the unclear results and time consuming in the sparse representation of image superresolution reconstruction with single redundant dictionary， a single image superresolution reconstruction method based on multidictionary learning and image patches mapping was proposed. In the framework of the traditional sparse representation， firstly the gradient structure information of local image patches was explored， and a large number of training image patches were clustered into several groups by their gradient angles， from those clustered patches the corresponding dictionary pairs were learned. And then the mapping function was computed from low resolution patch to high resolution patch in each clustered group via learned dictionary pairs with the idea of neighbor embedding. Finally the reconstruction process was reduced to a projection of each input patch into the high resolution space by multiplying with the corresponding precomputed mapping function， which improved the images quality with less running time. The experimental results show that the proposed method improves the visual quality significantly， and increases the PSNR （Peak SignaltoNoise Ratio） at least 0.4dB compared with the anchored neighborhood regression algorithm. 　　Key words：sparse representation； image patches classification； multidictionary learning； mapping function； neighbor embedding 　　0 引言 　　在医学诊断、卫星遥感、视频监控等数字成像领域中，都需要图像有较高的分辨率来为人的研究工作提供帮助。但是由于电子成像设备、环境以及传感器制造成本制约等因素的影响，实际获得的图像分辨率难以满足实际的应用需求， 因此通过软件技术手段，将输入的单幅或多幅低分辨率（Low Resolution， LR）图像恢复出相同场景的高分辨率（High Resolution， HR）图像的方法叫图像超