基于学习的超分辨率算法利用样本先验信息重建高分辨率图像,在遥感、刑侦和医学图像领域有着广泛应用。论文分析了前沿的基于稀疏表达的图像超分辨率算法,实现了该算法功能,为了便于基于稀疏表达超分辨率算法的应用,论文设计并实现了基于对话框和参数调节控件的图像超分辨率算法框架,实验结果表明论文实现的算法框架具有良好的可用性和拓展性。
The method of learning-based image super-resolution uses the prior of samples to rebuild the high-resolution image and has been extensively used in remote sensing, criminal investigation and medical image. This paper analyzes the lat- est image super-resolution method based on sparse representation and implemented this algorithm. For the better application of this method, this paper designs an algorithm framework based on dialog box and parameter control widgets. The result of experiments show that this framework has better availability and expansibility.