Multi-objects tracking and data association have re-
ceived considerable attention in the field of computer
vision, mobile robotic, autonomous systems and intel-
ligent transportation system (ITS). It has to deal with
difficulties existing in single object tracking, such as
changing appearances, non-rigid motion, dynamic illu-
mination and occlusion, as well as the problems related
to multiple objects tracking including inter-object oc-
clusion, multiple object confusion. To be effective, any
proposed method for multi-objects tracking has to meet
several stringent requirements:
• automatic segmentation of each moving object,
from the background, and from other objects, so
that all objects are detected,
• robust operation under a wide range of real-world
conditions, i.e. congestion, partial occlusions of
objects,
• robust operation in a wide variety of lighting con-
ditions, shadow, sunny, etcetera.
Even though a number of tracking methods have been
introduced in the literature, many of these criteria still
cannot be met. In our research work, we describe a
novel object tracking technique in color video sequen-
ces, with application to multi-object tracking in crow-
ded scenes. The proposed paradigm integrates object
detection into the object tracking process and provides
a robust tracking framework under ambiguity condi-
tions. In order to reduce the computational complex-
ity and to increase the robustness, we use a trisec-
tional structure. I.e., firstly it distinguishes between
real world objects, secondly extracts image features like
motion blobs and colour patches and thirdly abstracts
objects like meta-objects that shall denote real world
objects.

Through such a tight integration of the motion blobs
and colour patches, as well as the global optimization
of object trajectories, we have accomplished not only
robust and efficient multi-object tracking, but also the
ability to deal with merging/splitting of objects, irreg-
ular object motions, changing appearances, etc. which
are the challenging problems for the most traditional
tracking methods. For solving the problems of the
fluctuation detection and dealings with object interac-
tions, a data association step is suggested in further
step with data exclusion, data allocation and data ad-
ministration. The efficiency of the suggested technique
for multi-objects detection and tracking is demonstra-
ted and published in several papers on the basis of anal-
ysis of strongly disturbed real image sequences. (A. Al-
Hamadi -18709, S. Pathan -064, R. Niese -483)