Object Tracking and Position Determination: PG Scholar, E &C Dept, DBIT, Bangalore, Karnataka, India

OBJECT TRACKING AND POSITION DETERMINATION
DABBARU MURULI
PG Scholar, E &C Dept, DBIT, Bangalore, Karnataka, India
SHARAN BASAPPA
Asst. Professor, E & C dept, DBIT, Bangalore, Karnataka, India
ABSTRACT- Object tracking algorithm proposed based on color feature extraction, developed
algorithm implemented on MATLAB to track a Red, Green and Blue colored objects in a video
frame which is captured continuously from a live webcam, algorithm tested under different
circumstances and able to detect, track the movements of the particular object and determine its
position in the incoming frames.
KEYWORDS: Object Tracking, Feature Extraction, Color Feature, Position Determination
INTRODUCTION
The extraction of region of interest (ROI) in a real time environment is one of the difficult tasks;
there is a lot a computer can do in this situation. Mainly there are two approaches to perform this
operation i.e. pattern recognition and pixel by pixel computations. The first approach requires
more number of computations whereas second needs only few computations, the only drawback is
they are more sensitive to the noise and requires a static environment to function. Object tracking
is a process of acquiring images using smart cameras in an area under surveillance and recognizing
the region of interest i.e. Object. Further proceeding is to track the movements of the objects in a
sequence of frames. The images acquired from video may have some noise which should be
cleared using filters during pre-processing. Noise in the images can be removed using spatial
filtering techniques which improve the image quality. The filtering process applied considering the
neighboring pixels and scanning each element by rows and columns.
METHODOLOGY
The frames acquired from video input is converted into grey scale image and binary image
respectively after extracting only red components in the frames as the targeted region of interest
here is red object. Further taking the X-axis and Y-axis of the frame it is divided into 16
quadrants. Using the conditional status check the red object is determined in which quadrant it is in
the present position, as the video is continuously running if the object position changed then the
quadrant values is updated
Segmentation
The obtained digital image divided into several regions to find the objects and boundaries in the
image. Basically each pixel is unique in terms of color, intensity etc. These unique features are
used for the processing. The major part in object tracking is carried over here by finding region of
interest based on its feature here the specific feature used for the tracking is color. Extracting the
object using the specific feature in a sequence of incoming frames called object detection.
Algorithm developed such that the object in each frame can be found.
Tracking
Extracting the object using the specific feature in a sequence of incoming frames called object
detection. Algorithm developed such that the object in each frame can be found, Tracking can be
defined as finding a moving object in the incoming frames. This work is done by calculating the
position values of each pixel in the frames, with these values the position of entire object can be
achieved
IMPLEMENTATION
The block diagram of the proposed algorithm to implement on MATLAB is as shown below; the
major parts of the process are video input, pre-processing, extracting region of interest, tracking
and position determination.
Fig.1 Block Diagram
Description of main blocks

Video Input: Snapshots are obtained using image acquisition toolbox in the MATLAB from
webcam which is connected to the computer.
Pre-Processing: The incoming frames are tricolor or RGB Images, The RGB components are
individually extracted, and frames are converted into Grey scale image and binary images which
are convenient for the processing, in pre-processing the images are also filtered using built in
filters in MATLAB.
Extract Region of Interest: Segmentation performed on the converted and filtered binary images
to extract the region of interest i.e. red colored object here. All the connected red colored
components are recognized and extracted from the incoming frames.
Dilation is a basic operation in morphology. Initially developed for binary images, further
developed for grayscale images, and then to lattices. This creates a structured element for finding
and expanding the shapes present in the given input image
Let E be a Euclidean space or an integer grid, A is a binary image in E, and B a structuring
element.
The dilation can also be obtained by:
212
Erosion is one of the operations in morphological image processing. Erosion was developed
for binary images, further extended to grayscale, and to lattices
E is a Euclidean space or an integer grid, and A a binary image in E. The erosion of the binary
image A by the structuring element B can also be obtained by:
Tracking: Tracking is achieved based on recognizing the connected red components continuously
in the real time incoming video frames, various features and steps followed in algorithm to keep a
track of changes in the video frames.
Determine Object Position: Once after finding the red colored object in the image the X-axis and
Y-axis of the frame taken into consideration and divided the image into 16 quadrants start from 0
to 15th, algorithms developed works on to find out in which particular quadrant the connected red
components i.e. targeted object to track, the position of the object displayed on the computer
screen in MATLAB/GUI
PROPOSED ALGORITHM FLOWCHART
Fig. 2 Proposed Algorithm Flowchart
213
There are several algorithms to detect an object in the video frames, the above algorithm explains
step by step processing on object tracking and determining its position. Every algorithms
developed will have strengths and weakness, the algorithm which is used here can be help full for
further developments in the field of tracking.
APPLICATION DOMAIN
I. Automated video surveillance: Any moving objects in an area under surveillance can be
identified and report the doubtful objects.
II. Robot vision: The steering system can be designed to avoid collision and improve its
target based movements.
III. Traffic monitoring: The vehicles which violate traffic rules can be identified with
recorded videos by installing webcams on highways through which accidents can be
controlled.
IV. Animation: Object tracking algorithm developed here can also be extended to animation.
V. Human computer interaction: here, live video processing provides a very good
interaction of human and processing unit which helps to develop an application based
system.
LIMITATIONS
Tracking objects can be complex due to:

I. Dynamic environment: Environmental conditions are most important in image
acquisition, changes in light intensity can affect the tracking
II. Camera motion: If the camera in motion the incoming frames cannot be stable which
makes the algorithm difficult to track the object continuously
RESULTS
The work carried over real time video with proposed algorithm implementation is as follows, here,
the snapshots show all three colored objects detected individually as well as all together and
finding out movements, determining the position in each frame.
Fig.3 No objects
214
Fig.4 Only RED Object in 10th Quadrant
Fig.5 Only Green Object in 0th Quadrant
215
Fig.6 Only Blue Object in 5th Quadrant
Fig.7 All Three Objects
216
Here, Red object in 13th Quadrant, Green Object in 14th Quadrant and Blue object in 2nd Quadrant,
All three Objects are position is determined at a time.
FUTURE WORK
Significant progress has been made in object tracking during the last few years. Several robust
trackers have been developed which can track objects in real-time in simple scenarios. However, it
is clear that the assumptions used to make the tracking problem tractable, for example, smoothness
of motion, minimal amount of occlusions, illumination constancy, high contrast with respect to
background, etc., are violated in many respective scenarios
CONCLUSION
In this paper an algorithm is proposed to track the real-time objects on the basis of using the colour
property of the objects as a feature. In the context, Morphological dilation and erosion operations
are used in image processing to reduce noise during the real-time object tracking. Multiple objects
with multiple colours can be tracked and find their position at a time. May the contents discussed
in this paper can give valuable insight into this important research topic and encourage new
research.
ACKNOWLEDGMENT
I wish to express gratitude to my guide Mr Sharan Basappa who has helped me in the preparation
of this paper.
REFERENCES
Compressive tracking via oversaturated sub-region classifiers qiuping zhu, jia yan, dexiang deng,
issn 1751-9632, accepted on 3rd april 2013 doi: 10.1049/iet-cvi.2012.0248
Visual tracking: an experimental survey:arnold w. m. smeulders, senior member, ieee, dung m.
chu, student member, ieee, vol. 36, no. 7, july 2014I.S. Jacobs and C.P. Bean, Fine particles,
thin films and exchange anisotropy, in Magnetism, vol. III, G.T. Rado and H. Suhl, Eds. New
York: Academic, 1963, pp. 271-350.
A directional-edge-based real-time object tracking system employing multiple candidate-location
generation pushe zhao, hongbo zhu, he li, and tadashi shibata, member, ieee, ieee transactions
on circuits and systems for video technology, vol. 23, no. 3, march 2013
W. Hu, X. Zhou, W. Li, W. Luo, X. Zhang, and S. Maybank, Active contour-based visual
tracking by integrating colors, shapes, and motions, IEEE Trans. Image Process., vol. 22, no. 5,
pp. 17781792, May. 2013.
E. Maggio and A. Cavallaro, Tracking by sampling trackers, in Proc. IEEE ICCV, Barcelona,
Spain, 2011, pp. 11951202.
X. Mei, H. Ling, Y. Wu, E. Blasch, and L. Bai, Minimum error bounded efficient l1 tracker with
occlusion detection, in Proc. IEEE CVPR, Providence, RI, USA, 2011.
S. Hare, A. Saffari, and P. H. S. Torr, Struck: Structured output tracking with kernels, in Proc.
IEEE ICCV, Barcelona, Spain, 2011.
K. E. A. van de Sande, J. R. R. Uijlings, T. Gevers, and A. W. M. Smeulders, Segmentation as
selective search for object recognition, in Proc. ICCV, Barcelona, Spain, 2011.
217
Multiple-shot person re-identification by chromatic and epitomic analyses, PRL, vol. 33, no. 7,
pp. 898903, 2012. Z. Qigui and L. Bo, Search on automatic target tracking based on PTZ
system, in Proc. IEEE IASP, Hubei, China, 2011, pp. 192195.
A. Amato, M. G. Mozerov, A. D. Bagdanov, and J. Gonzalez, Accurate moving cast shadow
suppression based on local color constancy detection, IEEE Trans. Image Process., vol. 20, no.
10, pp. 29542966, Oct. 2011.
218

Object Tracking and Position Determination: PG Scholar, E &C Dept, DBIT, Bangalore, Karnataka, India

Uploaded by

Copyright:

Available Formats

Object Tracking and Position Determination: PG Scholar, E &C Dept, DBIT, Bangalore, Karnataka, India

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Object Tracking and Position Determination: PG Scholar, E &C Dept, DBIT, Bangalore, Karnataka, India

Uploaded by

Copyright:

Available Formats

OBJECT TRACKING AND POSITION DETERMINATION

KEYWORDS: Object Tracking, Feature Extraction, Color Feature, Position Determination

Fig.1 Block Diagram

Description of main blocks

The dilation can also be obtained by:

PROPOSED ALGORITHM FLOWCHART

Fig. 2 Proposed Algorithm Flowchart

Tracking objects can be complex due to:

Fig.5 Only Green Object in 0th Quadrant

Fig.7 All Three Objects

You might also like