IEEE Trans Image Process. 2015 Mar;24(3):846-61. doi: 10.1109/TIP.2015.2389629. Epub 2015 Jan 7.
Example learning-based superresolution (SR) algorithms show promise for restoring a high-resolution (HR) image from a single low-resolution (LR) input. The most popular approaches, however, are either time- or space-intensive, which limits their practical applications in many resource-limited settings. In this paper, we propose a novel computationally efficient single image SR method that learns multiple linear mappings (MLM) to directly transform LR feature subspaces into HR subspaces. In particular, we first partition the large nonlinear feature space of LR images into a cluster of linear subspaces. Multiple LR subdictionaries are then learned, followed by inferring the corresponding HR subdictionaries based on the assumption that the LR-HR features share the same representation coefficients. We establish MLM from the input LR features to the desired HR outputs in order to achieve fast yet stable SR recovery. Furthermore, in order to suppress displeasing artifacts generated by the MLM-based method, we apply a fast nonlocal means algorithm to construct a simple yet effective similarity-based regularization term for SR enhancement. Experimental results indicate that our approach is both quantitatively and qualitatively superior to other application-oriented SR methods, while maintaining relatively low time and space complexity.
基于示例的超分辨率 (SR) 算法在从单个低分辨率 (LR) 输入中恢复高分辨率 (HR) 图像方面显示出很大的潜力。然而，最流行的方法要么是时间密集型的，要么是空间密集型的，这限制了它们在许多资源有限的环境中的实际应用。在本文中，我们提出了一种新颖的计算高效的单图像 SR 方法，该方法学习多个线性映射 (MLM)，以直接将 LR 特征子空间转换为 HR 子空间。具体来说，我们首先将 LR 图像的大型非线性特征空间划分为一组线性子空间。然后学习多个 LR 子字典，然后根据 LR-HR 特征共享相同表示系数的假设，基于推断相应的 HR 子字典。我们建立从输入 LR 特征到期望 HR 输出的 MLM，以实现快速而稳定的 SR 恢复。此外，为了抑制基于 MLM 的方法产生的令人不快的伪影，我们应用快速非局部均值算法来构建一个简单而有效的基于相似性的正则化项，以增强 SR。实验结果表明，我们的方法在定量和定性上都优于其他面向应用的 SR 方法，同时保持相对较低的时间和空间复杂度。
