Editing Computer vision (section)

==History==
In the late 1960s, computer vision began at universities that were pioneering [[artificial intelligence]]. It was meant to mimic the [[human visual system]] as a stepping stone to endowing robots with intelligent behavior.<ref name="Szeliski2010" /> In 1966, it was believed that this could be achieved through an undergraduate summer project,<ref>{{Cite book |last=Sejnowski |first=Terrence J. |title=The deep learning revolution |date=2018 |publisher=The MIT Press |isbn=978-0-262-03803-4 |location=Cambridge, Massachusetts London, England |pages=28}}</ref> by attaching a camera to a computer and having it "describe what it saw".<ref name="Seymour1966" /><ref name="Boden2006" />

What distinguished computer vision from the prevalent field of [[digital image processing]] at that time was a desire to extract [[three-dimensional]] structure from images with the goal of achieving full scene understanding. Studies in the 1970s formed the early foundations for many of the computer vision [[algorithm]]s that exist today, including [[Edge detection|extraction of edges]] from images, labeling of lines, non-polyhedral and [[Polyhedron model|polyhedral modeling]], representation of objects as interconnections of smaller structures, [[optical flow]], and [[motion estimation]].<ref name="Szeliski2010" />

The next decade saw studies based on more rigorous mathematical analysis and quantitative aspects of computer vision. These include the concept of [[Scale space|scale-space]], the inference of shape from various cues such as [[shading]], texture and focus, and [[Active contour model|contour models known as snakes]]. Researchers also realized that many of these mathematical concepts could be treated within the same optimization framework as [[Regularization (mathematics)|regularization]] and [[Markov random field]]s.<ref name="Kanade20122">{{cite book|url=https://books.google.com/books?id=e0jtBwAAQBAJ&pg=PA191|title=Three-Dimensional Machine Vision|date=6 December 2012|publisher=Springer Science & Business Media|isbn=978-1-4613-1981-8|author=Takeo Kanade}}</ref>
By the 1990s, some of the previous research topics became more active than others. Research in [[Projective geometry|projective]] [[3D reconstruction|3-D reconstructions]] led to better understanding of [[Camera resectioning|camera calibration]]. With the advent of optimization methods for camera calibration, it was realized that a lot of the ideas were already explored in [[bundle adjustment]] theory from the field of [[photogrammetry]]. This led to methods for sparse [[3D reconstruction from multiple images|3-D reconstructions of scenes from multiple images]]. Progress was made on the dense stereo [[correspondence problem]] and further multi-view stereo techniques. At the same time, [[Graph cuts in computer vision|variations of graph cut]] were used to solve [[image segmentation]]. This decade also marked the first time statistical learning techniques were used in practice to recognize faces in images (see [[Eigenface]]). Toward the end of the 1990s, a significant change came about with the increased interaction between the fields of [[Computer graphics (computer science)|computer graphics]] and computer vision. This included [[image-based rendering]], [[Morphing|image morphing]], view interpolation, [[Image stitching|panoramic image stitching]] and early [[Light field|light-field rendering]].<ref name="Szeliski2010" />

Recent work has seen the resurgence of [[Feature (computer vision)|feature]]-based methods used in conjunction with machine learning techniques and complex optimization frameworks.<ref name="Sebe2005">{{cite book|author1=Nicu Sebe|author2=Ira Cohen|author3=Ashutosh Garg|author4=Thomas S. Huang|title=Machine Learning in Computer Vision|url=https://books.google.com/books?id=lemw2Rhr_PEC|date=3 June 2005|publisher=Springer Science & Business Media|isbn=978-1-4020-3274-5}}</ref><ref name="Freeman2008">{{cite journal|author1=William Freeman|author2=Pietro Perona|author3=Bernhard Scholkopf|title=Guest Editorial: Machine Learning for Computer Vision|journal=International Journal of Computer Vision|date=2008|volume=77|issue=1|pages=1|doi=10.1007/s11263-008-0127-7|issn=1573-1405|doi-access=free|hdl=21.11116/0000-0003-30FB-C|hdl-access=free}}</ref> 
The advancement of [[Deep learning|Deep Learning]] techniques has brought further life to the field of computer vision. The accuracy of deep learning algorithms on several benchmark computer vision data sets for tasks ranging from classification,<ref name="NatureBengio">{{cite journal |last2=Bengio |first2=Yoshua |last1=LeCun |first1= Yann| last3=Hinton | first3= Geoffrey|s2cid=3074096 |year=2015 |title=Deep Learning |journal=Nature |volume=521 |issue=7553 |pages=436–444 |doi=10.1038/nature14539 |pmid=26017442|bibcode=2015Natur.521..436L |url=https://hal.science/hal-04206682/file/Lecun2015.pdf }}</ref> segmentation and optical flow has surpassed prior methods.<ref>{{cite arXiv |eprint=1612.01925 |last1=Ilg |first1=Eddy |last2=Mayer |first2=Nikolaus |last3=Saikia |first3=Tonmoy |last4=Keuper |first4=Margret |last5=Dosovitskiy |first5=Alexey |last6=Brox |first6=Thomas |title=FlowNet 2.0: Evolution of Optical Flow Estimation with Deep Networks |date=2016 |class=cs.CV }}</ref><ref>{{Cite journal |title=A Survey of Deep Learning-Based Object Detection |year=2019 |doi=10.1109/ACCESS.2019.2939201 |arxiv=1907.09408 |last1=Jiao |first1=Licheng |last2=Zhang |first2=Fan |last3=Liu |first3=Fang |last4=Yang |first4=Shuyuan |last5=Li |first5=Lingling |last6=Feng |first6=Zhixi |last7=Qu |first7=Rong |journal=IEEE Access |volume=7 |pages=128837–128868 |bibcode=2019IEEEA...7l8837J |s2cid=198147317 }}</ref>