Robust Local Optical Flow Libary Documentation  V 1.2
 All Classes Files Functions Variables Enumerations Enumerator Macros Pages
mex_RLOF.m
1 % MEX_RLOF is a wrapper to the RLOF library that contains a set of sparse
2 % motion estimation methods based on the local optical flow paradigm.
3 %
4 % Usage:
5 %
6 % POINTLIST2 = mex_RLOF(IMAGE1, IMAGE1, POINTLIST1, PARAMSTRUCT);
7 %
8 % Arguments IMAGE1 and IMAGE1 could be gray scaled images (2-dimensional
9 % arrays) or rgb images (3-dimensional arrays)
10 % They are required to have the same dimension and they must be doubles.
11 %
12 % POINTLIST1 is the list of points to track (double 2xn array).
13 % With n the number of points in the point list.
14 %
15 % POINTLIST2 is the list of tracked points corresponding to "pointlist"
16 % (double 2xn array).
17 %
18 % PARAMSTRUCT specifies one or more of the following name/value pairs. Only
19 % the 'method' is mandatory.
20 %
21 % Name Value
22 % 'method' A string specifying the applied motion estimation
23 % method. The value can be one of the following:
24 % 'PLK' - Pyramidal Lucas Kanade provided by
25 % OpenCV.
26 % 'BEPLK' - Pyramidal Lucas Kanade using
27 % bilinear equations, see
28 % Senst et al. 2013.
29 % 'CB_PLK' - Pyramidal Lucas Kanade with
30 % adaptive support region, see
31 % Senst et al. 2014.
32 % 'CB_BEPLK' - Pyramidal Lucas Kanade using
33 % bilinear equations and an adaptive
34 % support region, see
35 % Senst et al. 2014.
36 %
37 % 'RLOF' - Robust Local Optical Flow, see
38 % Senst et al. 2011/2012.
39 % 'BERLOF' - Robust Local Optical Flow using
40 % bilinear equations, see
41 % Senst et al. 2013.
42 % 'CB_RLOF' - Robust Local Optical Flow with
43 % adaptive support region, see
44 % Senst et al. 2014.
45 % 'CB_BERLOF' - Robust Local Optical Flow using
46 % bilinear equations and an adaptive
47 % support region, see
48 % Senst et al. 2014. Full paper
49 % references are given below.
50 %
51 % 'noLevels' A positive number specifying the number of
52 % levels used for the gaussian image pyramid.
53 %
54 % 'noIterations' A positive number specifying the number of
55 % levels used for the gaussian image pyramid.
56 %
57 % 'winSize' A positive number specifying the size of the
58 % support region. Using RLOF or CB derivates
59 % winSize specify the upper bound of the support
60 % region size.
61 %
62 % 'smallWinSize' Used only for RLOF and CB derivates. A positive
63 % number specifying the lower bound of the support
64 % region.
65 %
66 % 'CBThreshold' Used only for CB derivates. A positive number
67 % specifying the pixel value threshold \tau, see
68 % Senst et al. 2014
69 %
70 % 'norm' A vector [V1 V2] specifying the shrinked Hampel
71 % norm \sigma_1 and \sigma_2, see
72 % Senst et al. 2014.
73 %
74 % 'options' A string specifying the debug options. The value
75 % can be one of the following:
76 % 'PerformTest' - A accuracy check with pre-
77 % estimated motion vectors for
78 % the ErnstReuter image pair
79 % will be done.
80 % 'PrintParameter' - Displays the parameter of the
81 % respective algorithm.
82 % Tobias Senst
83 % TU-Berlin
84 % 17. Nov, 2014
85 %
86 % COPYRIGHT:
87 %
88 % This file is the property of the author and Communication Systems Group,
89 % Technische Universität Berlin. All rights reserved.
90 %
91 % It may not be publicly disclosed, distributed, used, copied or modified
92 % without prior written authorization by a representative of
93 % Communication Systems Group, Technische Universität Berlin or the author.
94 % Any modified version of this document needs to contain this header.
95 %
96 % THERMS IF USAGE:
97 % PERSONAL, NON-COMMERCIAL or ACADEMIC USAGE:
98 % You are free to use this software for whatever you like. If you use this
99 % algorithm for a scientific publication, please cite the one of the
100 % following paper:
101 %
102 % @INPROCEEDINGS{ICIPSenst2014,
103 % AUTHOR = {Tobias Senst and Thilo Borgmann and Ivo Keller and Thomas
104 % Sikora},
105 % TITLE = {Cross based Robust Local Optical Flow},
106 % BOOKTITLE = {21th IEEE International Conference on Image Processing},
107 % YEAR = {2014},
108 % MONTH = okt,
109 % PAGES = {1967--1971},
110 % ADDRESS = {Paris, France},
111 % }
112 %
113 % @INPROCEEDINGS{ICIPSenst2013,
114 % AUTHOR = {Tobias Senst and Jonas Geistert and Ivo Keller and Thomas
115 % Sikora},
116 % TITLE = {Robust Local Optical Flow Estimation using Bilinear Equations
117 % for Sparse Motion Estimation},
118 % BOOKTITLE = {20th IEEE International Conference on Image Processing},
119 % YEAR = {2013},
120 % MONTH = sep,
121 % PAGES = {2499--2503},
122 % ADDRESS = {Melbourne, Australia},
123 % DOI = {10.1109/ICIP.2013.6738515},
124 % }
125 %
126 % @ARTICLE{TCSVTSenst2012,
127 % AUTHOR = {Tobias Senst and Volker Eiselein and Thomas Sikora},
128 % TITLE = {Robust Local Optical Flow for Feature Tracking},
129 % JOURNAL = {IEEE Transactions on Circuits and Systems for Video
130 % Technology},
131 % YEAR = {2012},
132 % MONTH = sep,
133 % PAGES = {1377--1387},
134 % VOLUME = {22},
135 % NUMBER = {9},
136 % DOI = {10.1109/TCSVT.2012.2202070}
137 % }
138 %
139 % @INPROCEEDINGS{WACVSenst2011,
140 % AUTHOR = {Tobias Senst and Volker Eiselein and Rubén Heras Evangelio and
141 % and Thomas Sikora},
142 % TITLE = {Robust Modified L2 Local Optical Flow Estimation and Feature
143 % Tracking},
144 % BOOKTITLE = {IEEE Workshop on Motion and Video Computing},
145 % YEAR = {2011},
146 % MONTH = jan,
147 % EDITOR = {Eric Mortensen},
148 % PAGES = {685--690},
149 % ADDRESS = {Kona, USA},
150 % DOI = {10.1109/WACV.2011.5711571},
151 % }
152 %
153 % COMMERCIAL USAGE:
154 % It is not allowed to use any content of this package for any commercial
155 % use or any advertisement for upcoming commercial products. If you want to
156 % use any content for such a purpose please contact:
157 % Prof. Dr.-Ing. Thomas Sikora <sikora@nue.tu-berlin.de>.
158 %
159 % WARRANTIES:
160 %
161 % Software provided by Technische Universität Berlin with this document is
162 % provided "AS IS" and any express of implied warranties including, but
163 % not limited to, the implied warranties of merchantability and fitness
164 % for a particular purpose are disclaimed.
165 % In no event shall the author or contributors be liable for any direct,
166 % indirect, incidental, special, exemplary, or consequential damages
167 % (including, but not limited to, procurement of substitute goods or
168 % services, loss of use, data, or profits or business interruption) caused
169 % in any way out of the use of this software, even if advised of the
170 % possibility of such damage.