图像的匹配调研从两个方向进行了先广的搜索，进行了初步的分析和资料分类，第一个方向是slam相关技术比较重要的一个话题就是智能体的位姿估计，这个估计过程中图像匹配是一个重要方法，在这个领域里这个话题叫做视觉里程计；另一个方向就是直接从传统的图像处理社区找相关方法。

视觉里程计

• 综述
  • 《Evolution of Visual Odometry Techniques》2018
  • 《视觉slam十四讲》第七讲、第八讲
  • 《Computer vision for autonomous vehicles: Problems, datasets and state-of-the-art》* 2017
  • 《Visual SLAM algorithms: a survey from 2010 to 2016》
• 基于几何方法（基于摄像头的几何特征计算）
  • feature-based
    • five-point algorithm + bundle ajustment refine
      • 《Bundle adjustment rules》 2006
      • Bundle adjustment介绍：https://blog.csdn.net/OptSolution/article/details/64442962
    • feature-based VO在火星车上的使用
      • 《Two years of visual odometry on the mars exploration rovers》2007
    • three-plus-one algorithm by using vanishing point or gravitational vector
      • 《Two efficient solutions for visual odometry using directional correspondence》2012
    • Epipolar search
      • 《Parallel, real-time monocular visual odometry》* 2013
      • 利用单目相机移动后对同一个标记点的两次观测，推导出对极约束，进而得到essential矩阵（八点法）
      • 再根据essential矩阵求得旋转矩阵R和偏移t*（svd）
      • 单应点
    • 多帧数据融合
      • 《Visual odometry by multi-frame feature integration》2013
    • 快速近似最近邻flann
    • 特征描述：
      • sift/surf/orb/brief/brisk
      • sift:《Distinctive Image Features from Scale-Invariant Keypoints》2004
      • surf:《surf speeded up robust features》2006
      • orb:《orb an efficient alternative to sift》 2011
    • 特征选择
      • reducing outlier
        • 《A temporally consistent grid-based visual odometry framework for multi-core architectures》2015
      • 对特征进行分类，只匹配部分特征
        • 《A Classification-Based Visual Odometry Approach》2016
      • 用惯性运动模型对特征进行筛选
        • 《Inertia constrained visual odometry for navigational applications》2017
        • 《A Framework for Fast and Robust Visual Odometry》* 2017
  • appearance-based
    • 不通过特征点，而是通过光度的变化来计算
    • region-based
      • 《Appearance-based monocular visual odometry for ground vehicles》* 2011
      • 《Correlation-based visual odometry for ground vehicles》* 2011
      • 《Adaptive-search template matching technique based on vehicle acceleration for monocular visual odometry system》2016
      • 《Robust Localization Using Eigenspace of Spinning-Images》* 2000
      • 《Accurate visual odometry from a rear parking camera》* 2011
      • fft
        • 《Visual vehicle egomotion estimation using the fourier-mellin transform》2007
        • 《Image-based memory for robot navigation using properties of omnidirectional images》2004
    • optical flow based
      • 《Robust motion estimation under varying illumination》2005
      • 《Autonomous navigation in three- dimensional urban environments using wide-field integration of optic flow》2010
      • 《A robust visual odometry and precipice detection system using consumer- grade monocular vision》2005
      • 《Fast visual odometry using intensity-assisted iterative closest point》* 2016
      • 稀疏光流（LK光流），稠密光流（HS光流）
  • hybrid-based
    • 《Dense monocular depth estimation in complex dynamic scenes》2016
    • 《SVO: Fast semi-direct monocular visual odometry》* 2014
• 基于非几何方法（机器学习方法训练模型计算）
  • 《Learning Visual Odometry with a Convolutional Network》2015
  • 《Deep image homography estimation》2016
  • 《Learning to relate images》2013
  • 《DeepVO: a deep learning approach for monocular visual odometry》2016
  • 《Deepvo: Towards end- to-end visual odometry with deep recurrent convolutional neural networks》2017
  • 《Predictive monocular odometry (PMO): What is possible without RANSAC and multiframe bundle adjustment?》* 2017
  • 《Unsupervised Learning of Monocular Depth Estimation and Visual Odometry with Deep Feature Reconstruction》* 2018
  • 《Flownet 2.0: Evolution of optical flow estimation with deep networks》2017
  • 《Flownet: Learning optical flow with convolutional networks》2015

图像配准

• 综述
  • 《Image Registration Techniques: A Survey》2017
• correlation based methods
  • 《Multimodality image registration by maximization of mutual information》1997
  • 《An FFT-Based Technique and Best-first Search for Image Registration》 2008
  • 《Image Registration Using Log Polar Transform and Fft Based Scale Invariant》 2014
  • 《A Fast and Robust Matching Framework for Multimodal Remote Sensing Image Registration》2018 (fft-based)
  • 《A Subpixel Registration Algorithm for Low PSNR Images》2018
• mutual information based methods
  • 《Alignment by maximization of mutual information》1997
  • 《Spline pyramids for inter-modal image registration using mutual information》1997
• wavelet based methods
  • 《Combination of feature-based and area- based image registration technique for high resolution remote sensing image》2007
• machine learning-based
  • 《VoxelMorph: A Learning Framework for Deformable Medical Image Registration》2018 https://github.com/voxelmorph/voxelmorph
  • 《End-to-End Unsupervised Deformable Image Registration with a Convolutional Neural Network》2017
  • 《Deep Learning based Inter-Modality Image Registration Supervised by Intra-Modality Similarity》2018
  • 《Quicksilver: Fast Predictive Image Registration – a Deep Learning Approach》2017
  • 《Real-time Deep Pose Estimation with Geodesic Loss for Image-to-Template Rigid Registration》2018
  • 《Elastic Registration of Medical Images With GANs》2018
  • 《Non-rigid image registration using fully convolutional networks with deep self-supervision》
• 《A matrix-free approach to parallel and memory-efficient deformable image registration》*2018
• 《A Fast and Accurate Nonlinear Spectral Method for Image Recognition and Registration》2006
• 《Optimized hierarchical block matching for fast and accurate image registration》
• 《Fast parallel image registration on CPU and GPU for diagnostic classification of Alzheimer’s disease》* 2014 https://github.com/SuperElastix/elastix https://github.com/SuperElastix/SimpleElastix
• 《Fast and Robust Symmetric Image Registration Based on Intensity and Spatial Information》 2018
• 《An Artificial Agent for Robust Image Registration》2017

联系我: