Доступ предоставлен для: Guest
Портал Begell Электронная Бибилиотека e-Книги Журналы Справочники и Сборники статей Коллекции
Journal of Flow Visualization and Image Processing
SJR: 0.161 SNIP: 0.312 CiteScore™: 0.1

ISSN Печать: 1065-3090
ISSN Онлайн: 1940-4336

Выпуски:
Том 27, 2020 Том 26, 2019 Том 25, 2018 Том 24, 2017 Том 23, 2016 Том 22, 2015 Том 21, 2014 Том 20, 2013 Том 19, 2012 Том 18, 2011 Том 17, 2010 Том 16, 2009 Том 15, 2008 Том 14, 2007 Том 13, 2006 Том 12, 2005 Том 11, 2004 Том 10, 2003 Том 9, 2002 Том 8, 2001 Том 7, 2000 Том 6, 1999 Том 5, 1998 Том 4, 1997 Том 3, 1996 Том 2, 1995 Том 1, 1993

Journal of Flow Visualization and Image Processing

DOI: 10.1615/JFlowVisImageProc.2017020118
pages 275-321

VISUAL TRACKING WITH METHODOLOGIES − A LITERATURE SURVEY

Xiang Xuezhi
College of Information & Communication Engineering, Harbin Engineering University, Harbin 150001, Heilongjiang, P.R. China
Syed Masroor Ali
College of Information & Communication Engineering, Harbin Engineering University, Harbin 150001, Heilongjiang, PR China

Краткое описание

Various algorithms used in computer vision are discussed here. The Kalman filter widely used in various computer vision applications includes object tracking, pattern recognition, etc. The Kalman filter was further improved to an extended Kalman filter with some additional features cited in this paper. Subsequently, the Particle filter as a filtering algorithm was developed with some distinct features. However, like Kalman filters, they opened a great way for tracking the state of a dynamic system for which one has a Bayesian model. The Particle filter was further modified to an improved particle filtering tracking algorithm for focusing on the problem of tracking a moving object in image sequences with complex background. Moreover, for somewhat similar type of applications, Optical flow was introduced. The term "Optical flow" is also used by roboticists, encompassing related techniques from image processing and control of navigation including motion detection, object segmentation, time-to-contact information, focus of expansion calculations, luminance, motion- compensated encoding, and stereo disparity measurement. It is very helpful for 2D motion and tracking. Considering Optical flow as a base, another flow technique known as Scene flow/Scene flow estimation is devised for 3D motion and tracking that is very helpful in state-of-the-art Visual Odometry and SLAM algorithms. This paper is a critical review/survey on the existing literature relevant to computer vision algorithms and their applications. Emphasis has been put on succinctly providing the knowledge regarding all previous and latest algorithms implemented in the field of computer vision.