基于SVM预分类学习的图像超分辨率重建算法

针对一般基于范例学习超分辨率重建算法的图像块误匹配和运算复杂度高等问题,提出了一种基于支持向量机预分类学习的算法。通过在匹配搜索前使用SVM筛选出与重建目标图像颜色特征相似的样本子库,保证了精确匹配搜索过程中样本块与输入低分辨率图像块内容之间的相关性,大大减少了误匹配现象,从而提高了图像重建质量。实验结果表明,算法的重建效果优于基于范例学习的算法,并在保证重建精度的前提下有效提高了算法运行速度。

Image super-resolution algorithm based on SVM pre-classified learning

Example-based super-resolution algorithm needed to run though the sample library with a high computing complexity. This method resulted in high calculation load and image degradation because of mis-matching. To resolve such problems, this paper proposed an algorithm based on SVM pre-classified learning. Before searching, it selected the subset of sample library similar to the color feature of object image, so as to ensure the content relevance between the sample patch and the input low-resolution image. In addition, the algorithm reduced the mis-matching greatly. The experimental results show the proposed algorithm has a better reconstruction performance than the example-based algorithm, which improves the program running speed in the precondition of accuracy.