Object Detection and Tracking Techniques Comparison
1. Mrs. K. V. Joshi
Asst. Professor (E&Tc Dept.)
SIT College of Engg
Yadrav-Ichalkaranji
2. Contents
Image Processing
What is an image
Digital Image Representation
Various Color Models
Object Detection
Object Tracking
Challenges in Object Detection and Tracking
Object Tracking Approaches
Applications of Object Detection and Tracking
Steps involved in Object Detection and Tracking
4. Image Processing
In imaging science, image processing is processing of
images using mathematical operations by using any form
of signal processing for which the input is an image, a
series of images, or a video, such as a photograph or video
frame; the output of image processing may be either an
image or a set of characteristics or parameters related to
the image.
5. What is an Image?
An image may be defined as a two-dimensional function ,
f(x,y), where x and y are spatial (plane) coordinates, and
the amplitude of f at any pair of coordinates(x, y) is called
the intensity or gray level of the image at that point.
When x, y, and the intensity values of f are all finite,
discrete quantities, we call the image a digital image. The
field of digital image processing refers to processing digital
images by means of a digital computer.
A digital image is composed of a finite number of
elements, each of which has a particular location and
value. These elements are called picture elements, image
elements, pels, and pixels. Pixel is the term used most
widely to denote the elements of a digital image.
6. Digital Image Representation
An image can be defined as a 2D signal that varies over
spatial co-ordinates x and y and can be written
mathematically as f(x,y)
Binary Image Equivalent Image contents in Binary Form
1 1 1 1 1
0 1 1 1 0
0 1 1 1 0
0 1 1 1 0
0 1 1 1 0
8. RGB Color Model
In the RGB model, an image
consists of three independent
image planes, one in each of the
primary colors: red, green and
blue. (The standard wavelengths
for the three primaries are as
shown in figure ). Specifying a
particular color is by specifying the
amount of each of the primary
components present.
Figure shows the geometry of the
RGB color model for specifying
colors using a Cartesian coordinate
system. The greyscale spectrum,
i.e. those colors made from equal
amounts of each primary, lies on
the line joining the black and
white
9. CMY Color Model
subtractive color model appropriate to absorption of
colors, for example due to pigments in paints. Whereas
the RGB model asks what is added to black to get a
particular color, the CMY model asks what is subtracted
from white. In this case, the primaries are cyan, magenta
and yellow, with red, green and blue as secondary colors
When a surface coated with cyan pigment is illuminated
by white light, no red light is reflected, and similarly for
magenta and green, and yellow and blue. The
relationship between the RGB and CMY models is given
by:
10. HSI Color Model
As mentioned above,
color may be specified
by the three quantities
hue, saturation and
intensity. This is the
HSI model, and the
entire space of colors
that may be specified
in this way is shown in
figure.
11. YIQ Color Model
The YIQ (luminance-in-phase-quadrature) model
is a recording of RGB for color television, and is a
very important model for color image processing.
The conversion from RGB to YIQ is given by:
𝑌
𝐼
𝑄
=
0.299 0.587 0.114
0.596 −0.275 −0.321
0.212 −0.523 0.311
𝑅
𝐺
𝐵
12. HSV Color Model
HSV stands for hue, saturation,
and value
HSV is the most common cylindrical
co-ordinate representation of points in
an RGB color model. This
representation rearranges the geometry
of RGB in an attempt to be more
intuitive and perceptually relevant than
the Cartesian (cube) representation.
Developed in the 1970s for computer
graphics applications, HSV is used
today in color pickers, in image
editing software, and less commonly in
image analysis and computer vision.
13. Video Processing
A video can be considered as a collection of images
indexed by time.
Video processing is an extension of image processing.
Digital video comprises a series of orthogonal
bitmap digital images displayed in rapid succession at
a constant rate. In the context of video these images
are called frames. We measure the rate at which
frames are displayed in frames per second (FPS).
14. Object detection
Object detection is a method of following an object
through successive image frames to determine its
relative movement with respect to other frames.
Object detection in videos involves verifying the
presence of an object in image sequences and possibly
locating it precisely for recognition
15. Object Tracking
Object tracking is a process of segmenting a region of
interest from a video and keep tracking of the motion,
position and also the occlusion.
The tracking is performed by monitoring objects’
temporal and spatial changes during the video
sequence, also its presence, position, size, shape, etc.
16. Challenges in Object Detection and
Tracking
The detection and tracking of objects in video can be complex
due to:
1. Loss of information caused by projection of the 3D world on
a 2D image,
2. Noise in images,
3. The complex object motion,
4. Non-rigid or the articulated nature of the objects,
5. Partial and full object occlusions,
6. Scene illumination changes
7. Complex object shapes,
8. The real time processing requirements.
17. Object Tracking Approaches
Based on the above information there are three basic
approaches for object tracking,
Point tracking approach.
Kernel tracking approach.
Silhouette tracking approach.
18. Video Surveillance
Robot Vision
Traffic Monitoring
Video Inpainting (Video Inpainting Refers To A Field Of Computer
Vision That Aims To Remove Objects Or Restore Missing Or Tainted
Regions Present In A Video Sequence By Utilizing Spatial And
Temporal Information From Neighboring Scenes.)
Animation
Surveillance/Monitoring Applications Surveillance/Monitoring
Applications
Security Cameras
Traffic Monitoring
Face Detection
People Counting
Vehicle Detection
Applications of Object Detection
and Tracking
19. Online Images
Security
Manufacturing Industry
Control Applications
Object Avoidance Object Avoidance
Automatic Guidance
Head Tracking for Video Head Tracking for Video
Conferencing
Many intelligent video analysis systems are intelligent
video analysis systems are based on motion detection and
tracking
20.
Input Video
Extract frames from
input video
Object Detection
Object Representation
Object Tracking
Moving Object
Steps involved in Object Detection
and Tracking:
21. CAMShift
The Camshift algorithm is improved by Meanshift
algorithm, and it is called Continuously Adaptive
Mean Shift algorithm.
The basic idea of Camshift is to use the Meanshift to
all frames of the video image, and take the result of
previous frame as initial value to the next frame, so
iteration.
22. Steps Of Camshift Algorithm
The Camshift algorithm can be summarized in the following
steps:
(1) Initialize the search window.
(2) Calculate the color probability distribution of the search
window (back projection).
(3) Run Meanshift algorithm, obtain the new size and
position of the search window.
(4) Re-initialize the size and position of the search window
in the next frame of video image by using the value that
calculated in step (3), and jump to step (2) to proceed.
23. Particle Filter
Particle filter is a filtering method based on Monte
Carlo and recursive Bayesian estimation.
The particle filters, also known as condensation filter
and they are suboptimal filters.
The core idea is that density distribution is present
using random sampling particles. There is no
restriction to the state vector to deal with nonlinear
and non-Gaussian problem, and it is the most general
Bayesian approach.
25. For Noisy Video
The three types of noise such as salt & pepper,
Gaussian and Speckle noise was inserted into the
original video and the algorithm was tested.
Salt & pepper noise
29. For Noisy Video
The three types of noise such as salt & pepper,
Gaussian and Speckle noise was inserted into the
original video and the algorithm was tested
33. Intensity Consideration for Both
Algorithms
Table 1- Nomenclature of videos according to intensity
variations
Video Change in intensity
A1, B1 5%
A2, B2 10%
A3, B3 15%
A4, B4 20%
A5, B5 25%
35. Discussion
As noise is introduced in video, elapsed time increases
for Particle Filter algorithm, but by using CAMShift
algorithm the elapsed time decreases.
For both algorithms Object detection and tracking is
better in case of Salt and Pepper noise than that of
Gaussian and Speckle Noise.
For both the videos, as the intensity of original video is
changed, elapsed time decreases and track loss rate
increases.
36. REFERENCES
A. Yilmaz, O. Javed, M. Shah, “Object Tracking: A Survey”, ACM Comput. Surv. 2006; 38(4):13.
C. Manisha, S. Amudha, G. Vinaya, “Object Detection and Tracking in Video Sequences”, ACEEE Int. J. on Signal & Image Processing,
Vol. 03, No. 01, Jan 2012, pp: 36-41
Sakharkar S. Kamble s. Khobragade A. “Particle Filter Based Object Tracking International Journal on Recent and Innovation Trends in
Computing and Communication”, Volume: 3 Issue: 2 ISSN: 2321-8169 155– 158
Jog A, Halbe S, “Multiple Object Tracking using CAMShift Algorithm in OpenCV”, IOSR Journal of VLSI and Signal Processing
(IOSR-JVSP), ISSN: 2319 – 4200, ISBN No. : 2319 – 4197 Volume 1, Issue 2 (Sep-Oct. 2012), PP 41-46
Gustafsson F, Gunnarsson F, Bergman N, Forsesell U, Jansson J, Karlsson R, Nordlund P,
“Particle Filters for Positioning, Navigation and Tracking”, IEEE Transactions on Signal Processing, Special issue on Monte Carlo methods for
statistical signal processing, pp 1-13
Fang-I Chu, “From Baysian to Particle Filter”, A Thesis Presented to the Faculty of San Fransisco State University, August 2009,pp 1-112
Rao G. M., Dr. Satyanarayana Ch, “Visual Object Target Tracking Using Particle Filter: A Survey”, I. J. Image, Graphics and Signal Processing
2013, pp 57-71
Guan Y, Chen X, Wu Y, Yang D, “An Improved Particle Filter Approach for Real-Time Pedestrian Tracking in Surveillance Video”, 2013
International Conference on Information Science and Technology Application(ICISTA-13), pp 173-177
Coskun M, Unal S, “Implementation of Tracking of a Moving Object Based on CAMShift Approach with a UAV”, ELSEVIER, 9th International
Conference Interdisciplinarity in Engineering, INTER-ENG 2015, 8-9 October 2015,Tirgu-Mures, Romania, pp 556-561
Mahajan S, Kurve C, “Video Object Detection Using Particle Filter”, International Journal of Computer Network and Security(IJCNS), Vol 4-
No. 1 Jan-Mar 2012, ISSN:0975-8283 PP 24-27
Chen X, Wu H, Xiang Li, Luo X, Qui T, “Real-time Visual Object Tracking via CAMShift-Based Robust Framework”, International Journal of
Fuzzy Systems, Vol 14, No 2, June 2012, pp 262-269
Mohammed A, Morris T, “An Improved CAMShift Algorithm for Object Detection and Extraction”, IT CoNvergence PRActice(INPRA),
Volume 2, Number 2, pp 55-65