Image quality enhancement in arbitrary viewpoint image synthesis by the sparse sampling

传统的虚拟视图图像合成方法存在的主要问题是难以精确地补全合成图像中的缺失像素，需要大量的数据存储和计算量。在本研究中，我们提出了利用稀疏采样和压缩感知技术来克服这些困难的方法。实验结果从定性和定量上证明了所提方法的有效性

Major problems of traditional methods to synthesize virtual-view images are that it is difficult to precisely complete missing pixels in synthesized images and that it requires huge amount of data storage and computational efforts. In this research, we propose methods to overcome these difficulties by using techniques of the sparse sampling and the compressed sensing. Experimental results show the validity of the proposed methods qualitatively and quantitatively