Visibility graphs for image processing
经过查询资料目前确认是：可见性图而不是可见度图
Department of Surgery and Cancer, Division of Computational and Systems Medicine, Imperial College London, London SW72AZ (United Kingdom)
The Molecular Biology of Metabolism Laboratory,
The family of image visibility graphs (IVGs) have been recently introduced as simple algorithms by which scalar fields can be mapped into graphs.
Here we explore the usefulness of such operator in the scenario of image processing and image classification.
We demonstrate that the link architecture of the image visibility graphs encapsulates relevant information on the structure of the images and we explore their potential as image filters and compressors.
图像压缩流程：
Encoder：规范化—镜像填充—图像卷积—空间下采样—量化（就近取整）—code 编码输出
Modle: 高斯尺度混合GSM(控制量化系数分布)+noise
Decoder： 卷积—上采样（在子像素卷积的情况下）—去标准化—clip.
规范化：规范化以每个通道值的分布为中心，并确保其具有近似单位方差；
镜像填充：使得图像编码器输出具有和8倍下采样相同的空间范围；解决图像的边缘问题；
卷积和空间下采样（并将图像通道从64个增加到128个，卷积神经网络的内容）之后是三个剩余块，前两个剩余块都有两个含有128个filter的卷积层，最后一个剩余块卷积层之后对系数进行下采样。
量化：量化采用就近取整函数，且做了相应的改进，
Sub-pix卷积层:有卷积和系数重组构成，
Clip ：将图像灰度值控制在0-255之间。
GSM：高斯尺度混合模型
http://chiuki.github.io/android-shaders-filters/#/29
http://cdn.tutsplus.com/active/uploads/legacy/tuts/070_effectsTester/Preview/EffectsTester.html
We introduce several graph features, including the novel concept of Visibility Patches, and show through several examples that these features are highly informative, computationally efficient and universally applicable for general pattern recognition and image classification tasks.
