Tracking feature points in time-varying images using an opportunistic selection approach

doi:10.1016/0031-3203(89)90073-3

Pattern Recognition

Volume 22, Issue 3, 1989, Pages 247-256

https://doi.org/10.1016/0031-3203(89)90073-3 Get rights and content

Abstract

In tracking of feature points in time-varying images, one major issue is to identify the feature points at different times that represent the same physical object point. This process is often called the ‘correspondence problem’ and is an important research issue in both motion analysis and stereopsis. This paper describes a new approach toward the tracking of feature points in time-varying images by reasoning about the trajectories of feature points. The advantage of this approach is that the solution can be computed efficiently and the algorithm can potentially be supported by multiple processor computer. This approach is tested extensively on synthesized data. The statistics on the experiments strongly suggests the generality and the robustness of the approach.

References (11)

B.K.P. Horn et al.
Determining optical flow
Artif. Intell.
(1981)
S. Barnard et al.
Disparity analysis of images
IEEE Trans. Pattern Analysis Mach. Intell.
(1980)
R. Jain
Dynamic scene analysis using pixel-based processes
IEEE Comput.
(1981)
W. Chow et al.
Computer analysis of planar culvilinear moving images
IEEE Trans. Comput.
(1977)
J. Roach et al.
Determining the movement of objects from a sequence of images
IEEE Trans. Pattern Analysis Mach. Intell.
(1980)

There are more references available in the full text version of this article.

Cited by (33)

Sperm motility analysis system implemented on a hybrid architecture to produce an intelligent analyzer
2020, Informatics in Medicine Unlocked
Citation Excerpt :
Various tracking methods are proposed and applied in different CASA systems. Optical flows, matching feature points in two frames using a hypothesis testing method [29,30], opportunistic selection strategy [31], Pool's method [32], real-time spermatozoa tracing system (RSTS) [14], principal component analysis [33,34], watershed segmentation, and frame difference methods are some examples of techniques. These methods were compared previously [35].
Much research and analysis in biomedicine involve image and video inspection using microscopes. Presently, scientists are dissatisfied with manual observations and assessments, when objective and enhanced data can be obtained by applying new technologies (such as image and video inspection) to biomedical fields, such as sperm analysis. Computer Assisted Sperm Analysis (CASA) systems, developed in the late 1980s, constitute third-generation methods of sperm analysis. This study aimed to develop a standalone medical image and video analysis system that is reconfigurable, flexible, reliable, deterministic, and robust. It proposed a new sperm motility analysis system running on a dual core Central Processing Unit (CPU) + field programmable gate arrays (FPGA) platform, under a real-time operating system (RTOS), which is a step ahead of the third-generation CASA systems.
The system hardware and related sperm detection and tracking algorithms were the novelty of this work. The image processing functions mainly run on FPGA, image acquisition, and calculations run on CPU, parallel with FPGA.
The result is a much faster, reliable, reconfigurable, and compact intelligent analyzer system.
Our prototype system was applied to sperm motility analysis; however, other image processing systems can be applied to this architecture. Additionally, the proposed tracking method for sperm track determination is simple, effective, and does not exert a load on the system.
Gaussian and Laplacian of Gaussian weighting functions for robust feature based tracking
2005, Pattern Recognition Letters
Object tracking algorithms found extensively in the computer vision literature either are inhibited by various assumptions such as simplicity of motion and shape characteristics of objects or are overly sensitive to noise. We propose and successfully test two new weighting functions for a feature-based object-tracking algorithm to achieve superior performance in tracking motion of non-rigid objects under noisy conditions. We present the implications of using the weighting functions in real and synthetic image sequences to overcome the noise produced at acquisition source (charge coupled device—CCD), or in the background environment. We also present a mechanism for determining the optimal weighting function based on image parameters, more specifically the edge characteristics of objects in the image.
Identifying truck correspondence in multi-frame imagery
2005, Transportation Research Part C: Emerging Technologies
This paper develops a model for automated matching of vehicles between aerial photos. We exploit the overlap from frame-to-frame in a sequence of aerial photos. The problem, in essence, is to determine if vehicles A, B, C, … in one photo, match vehicles X, Y, and Z in a temporally lagged second photo. The method is specifically applied to large trucks (tractor–trailer combinations). If a match can be established, the trucks can be joined to form trajectories. The need for an automated algorithm is pressing in view of the likely stream of data from various new aerial platforms (video, tethered balloons, unmanned airborne data collection systems) and the development of tools to automate the extraction of trucks from these images. In the first step, manual air photo interpretation is used to determine the exact match of the real trucks. Digitized points in the images represent observations of unique trucks, each of which is seen a number of times. A distance matrix is computed between all pairs of points (trucks) on the network. Then, a linear program is used to select matching trucks. A total of 10 pairs of photos incorporating 85 potential matches are described. Reasonable results are found, especially for implied speed estimates, which can be compared to the actual speeds from the true matches. Challenges for future extensions to more complex situations are explained.
Motion tracking as a constrained optimization problem
2003, Pattern Recognition
In this paper we pose the problem of tracking of a varying number of points through an image sequence as a multi-objective optimization problem with additional hard constraints. One of the objectives is to find smooth tracks based on second-order motion characteristics optimized over several frames. The corresponding optimization algorithm we present is a sequential heuristic search algorithm that adequately prunes the search tree in such a way that its exponential order remains low. When the algorithm is compared to other tracking algorithms, it turns out that the proposed algorithm is easier to tune and generally more efficient and more accurate.
Establishing motion correspondence using extended temporal scope
2003, Artificial Intelligence
This paper addresses the motion correspondence problem: the problem of finding corresponding point measurements in an image sequence solely based on positional information. The motion correspondence problem is most difficult when the target points are densely moving. It becomes even harder when the point detection scheme is imperfect or when points are temporarily occluded. Available motion constraints should be exploited in order to rule out physically impossible assignments of measurements to point tracks. The performance can be further increased by deferring the correspondence decisions, that is, by examining whether the consequences of candidate correspondences lead to alternate and better solutions. In this paper, we concentrate on the latter by introducing a scheme that extends the temporal scope over which the correspondences are optimized. The consequent problem we are faced with is a multi-dimensional assignment problem, which is known to be NP-hard. To restrict the consequent increase in computation time, the candidate solutions are suitably ordered and then additional combined motion constraints are imposed. Experiments show the appropriateness of the proposed extension, both with respect to performance as well as computational aspects.
A relaxation algorithm for real-time multiple view 3D-tracking
2002, Image and Vision Computing
Citation Excerpt :
Typically isolated objects are tracked using a Kalman filter approach to predict and update object location estimates from observations [26,48]. Previous work, such as Refs. [17–19], focuses on the cost function definition according to motion smoothness or geometric constraints taking into account object occlusion and reappearance. Techniques for statistical data association have also been applied to motion correspondence [26,27].
In this paper we address the problem of reliable real-time 3D-tracking of multiple objects which are observed in multiple wide-baseline camera views. Establishing the spatio-temporal correspondence is a problem with combinatorial complexity in the number of objects and views. In addition vision-based tracking suffers from the ambiguities introduced by occlusion, clutter and irregular 3D motion. In this paper we present a discrete relaxation algorithm for reducing the intrinsic combinatorial complexity by pruning the decision tree based on unreliable prior information from independent 2D-tracking for each view. The algorithm improves the reliability of spatio-temporal correspondence by simultaneous optimisation over multiple views in the case where 2D-tracking in one or more views are ambiguous. Application to the 3D reconstruction of human movement, based on tracking of skin-coloured regions in three views, demonstrates considerable improvement in reliability and performance. Results demonstrate that the optimisation over multiple views gives correct 3D reconstruction and object labelling in the presence of incorrect 2D-tracking whilst maintaining real-time performance.

View all citing articles on Scopus

View full text

Tracking feature points in time-varying images using an opportunistic selection approach

Abstract

Artif. Intell.

Disparity analysis of images

IEEE Trans. Pattern Analysis Mach. Intell.

Dynamic scene analysis using pixel-based processes

IEEE Comput.

Computer analysis of planar culvilinear moving images

IEEE Trans. Comput.

Determining the movement of objects from a sequence of images

IEEE Trans. Pattern Analysis Mach. Intell.