zoukankan      html  css  js  c++  java
  • Super-Resolution Restoration of MISR Images Using the UCL MAGiGAN System 超分辨率恢复

    作者是伦敦大学学院Mullard空间科学实验室成像组,之前做过对火星图像的分辨率增强。

    文章用了许多的图像处理方法获得特征和高分辨率的中间结果,最后用一个生产对抗网络获得更好的高分辨率结果。

    用的数据是MISR多角度成像数据,225282个训练样本,输入275m分辨率(64*64),得到68.75m(256*256)的分辨率结果

    中间整个的流程和数据的处理都没怎么看懂

    过程:

    The MAGiGAN SRR system is based on the

    mutual shape adapted [2] features from accelerated segment test (MSA-FAST) [3] combined with

    convolutional neural network (CNN) [4] feature matching (see stage 2 in Section 2.2),

    adaptive least-squares correlation (ALSC) and

    region growing (Gotcha) [5] (see stage 3 in Section 2.2),

    partial differential equation (PDE)-based total variation (TV) regularization (GPT) [6,7] (see stage 4 in Section 2.2),

    support vector machine (SVM) and

    graph cut (GC)-based shadow modelling and removal [8] (see stage 1 in Section 2.2), and

    the generative adversarial network (GAN) [9] based super-resolution refinement method (see stage 5 in Section 2.2).

    ---------------- 坚持每天学习一点点
  • 相关阅读:
    23. Sum Root to Leaf Numbers
    22. Surrounded Regions
    21. Clone Graph
    19. Palindrome Partitioning && Palindrome Partitioning II (回文分割)
    18. Word Ladder && Word Ladder II
    14. Reverse Linked List II
    20. Candy && Gas Station
    16. Copy List with Random Pointer
    ubuntu 下建立桌面快捷方式
    java基础篇-jar打包
  • 原文地址:https://www.cnblogs.com/tccbj/p/10800083.html
Copyright © 2011-2022 走看看