光場資料收集（轉）
阿新 • • 發佈：2018-11-11
Links About Light Field
- 斯坦福大學光場資料庫：老牌斯坦福大學計算機圖形實驗室(Computer Graphics Laboratory)提供的，該資料庫所在網站還提供了光場的採集裝置，相機標定以及視覺化工具。
- 斯坦福大學計算機圖形實驗室的Marc Levoy教授做的動畫模擬，利於理解。Flash applets on some technical aspects of photography，這裡面詳細地介紹了相機的各種引數變化對應的光路圖的變化，強烈推薦。
- HCI光場資料集，千呼萬喚始出來，有好長一段時間這個資料集突然消失了（可能是在維護資料）。如今以新的面貌重現天日，真的讓人喜出望外。對於其資料集，HCI提供的解碼工具；這是要建立與Middlebury齊名的資料集（包括評價排名）的節奏啊！（ps:日後整理，很感興趣）
- Ravi Ramamoorthi教授主頁，一位計算成像，計算機視覺領域的大神，他所在組發表了很多高質量的文章，詳情可參考他的主頁。
- 光場實驗室網站，研究光場領域，隸屬於Ravi Ramamoorthi教授。這裡將是研究光場領域深度影象獲取，三維重建以及去除高光等領域研究者的福音。
- 光場相機的締造者以及Lytro公司的創始人Ren Ng。看這裡是其建立的Lytro公司的主頁；最初Lytro的最大賣點在於先拍照後對焦，可是買賬的人並不多；購買者多數是攝影愛好者以及科研機構。最近Lytro公司開始進軍VR以及AR領域，看了下其裝置，怎一個大字了得（Ps: 話說Ng為何去教書了？）
- Lytro論壇，這裡有關於Lytro公司以及光場相機最新的應用，時常關注不至於落後於時代，不至於被世界殘忍的拋棄。
- MATLAB光場工具包，這個工具包從事光場研究的科研人員的福利，後文中我將詳細的介紹這個工具包的使用方法。
- LytroMeltdown，一位布拉格大學(Charles University)的學生拆解Lytro 1.0的資料，如果你想對光場相機的內部結構有更加深入的瞭解的話，這個網址有豐富的介紹。（目前僅有Lytro一代的拆解並沒有ILLUM的拆解，原因是過於昂貴，作者買不起）
- cocolib光場工具套件，這個套件實際上是一個處理凸優化問題的庫，既可以用命令列操作也可Matlab介面操作。該庫實現了目前集中常用的演算法諸如： inverse problems，基於總變分最小化的影象分割以及向量多標記交易成本函式；當然最重要的還有對於光場影象的分析函式套件（基於HCI發表的深度估計論文）。
- 著名的Middlebury資料集：來提供了Benchmark(左右視角的紋理圖以及對應的GT深度影象)，各種演算法的效能對比， 在雙目深度估計領域屬於最為權威的評估標準。
- Deep learning: Depth Estimation，目前基於雙目，多目以及利用光場進行深度估計的演算法已有很多；有些研究者利用當前比較火熱的深度學習的策略對單幅影象進行深度影象的提取；我本來想跑跑人家程式碼，可以目前還沒時間搞這些。
- Computer Vision Laboratory - CVLAB
- 寬頻網數字媒體技術實驗室的Yebin Liu 劉燁斌主頁
- 深度圖+原始彩色影象轉化成多視角動態gif，戳這裡，這算是深度影象的一個小小的應用。
Light Field Resources
This is a (work in progress) repo for collecting links to data sets, source code, and other resources related to research on light fields for computer vision.For further information and interaction within the light field community, have a look at:
- Google Community for the Matlab Light Field Toolbox
- Light Field Forum
- Mailing List / General Light Field Vision Google Group
Background Information / General Light Field Information
- Wikipedia
- plenoptic.info provides some nice visualizations on how micro lens based plenoptic cameras work
- Todor Georgievs Website insights into plenoptic cameras. No longer updated (?)
- A Taxonomy and Evaluation of Dense Light Field Depth Estimation Algorithms paper with an in depth overview of depth estimation approaches for 4D light fields
- Light Fields and Computational Imaging early survey of the theory and practice of light field imaging
- please add more :)
Other Light Field Datasets
- The (New) Stanford Light Field Archive
- Stanford Lytro Light Field Archive
- MIT Synthetic Light Field Archive
- 4D Light Field Dataset (CVIA Konstanz & HCI Heidelberg)
- HCI 4D Light Field Dataset
- Lytro first generation dataset
- EPFL Light-Field Image Dataset
- Disney High Spatio-Angular Resolution Light Fields
- Light field Saliency Dataset (LFSD)
- LCAV-31 - A Dataset for Light Field Object Recognition
- A 4D Light-Field Dataset for Material Recognition
- Data for: Occlusion-aware depth estimation using light-field cameras
- DDFF 12-Scene 4.5D Lightfield-Depth Benchmark
- please add more :)
Tools
- Matlab Light Field Toolbox
- cocolib light field suite
- Geometric light field camera calibration toolbox
- Blender addon to create synthetic light field data sets
- please add more :)
Algorithm Source Code
- Accurate Depth Map Estimation from a Lenslet Light Field Camera (*LF)
- Occlusion-aware depth estimation using light-field cameras (*LF_OCC)
- Empirical Bayesian Light-Field Stereo Matching by Robust Pseudo Random Field Modeling (RPRF)
- Robust Depth Estimation for Light Field via Spinning Parallelogram Operator (SPO)
- EPINET: A Fully-Convolutional Neural Network using Epipolar Geometry for Depth from Light Field Images
- DeepMVS: Learning Multi-View Stereopsis
- Total variation-based dense depth from multicamera array
- Depth from Combining Defocus and Correspondence Using Light-Field Cameras
- Robust Light Field Depth Estimation for Noisy Scene with Occlusion
- Depth from Shading, Defocus, and Correspondence Using Light-Field Angular Coherence
- please add more :)
Where applicable, the short name in parentheses denotes the acronym used on the 4D light field benchmark.
Workshops & Tutorials
- 1st Workshop on Light Fields for Computer Vision @ ECCV 2014
- 2nd Workshop on Light Fields for Computer Vision @ CVPR 2017
- please add more :)
People / Labs
- CVIA, Computer Vision and Image Analysis, Uni Konstanz, Germany
- SCI, Stanford Computational Imaging, Stanford University, USA
- HCI, Heidelberg Collaboratory for Image Processing, Heidelberg University, Germany
- Home Page for Shuo Zhang, Beijingjiaotong
- Computer Vision and Image Analysis, Konstanz University
- Hae-Gon Jeon, Carnegie Mellon University
- Plenoptic Imaging, ACFR Marine,Sydney University
- Advanced Multimedia Processing Lab -- Projects -- The Self-Reconfigurable Camera Array
- Ting-Chun Wang's Homepage
- Pratul Srinivasan， EECS Department at UC Berkeley
Deep Learning Related
- Deep Learning in Matlab
- please add more :)
Others
Convert Latex to images
轉自https://www.vincentqin.tech/collections/,有補充