Video Surveillance systems
for Traffic Monitoring
Simeon Indupalli
1
Presentation Overview






Video surveillance systems.
Traffic monitoring issues.
Object tracking techniques.
Vehicle tracking strategies.
A real time system Explanation.
Future Work
2
what is video surveillance?

Present Implementations?



Advantages of video surveillance?



Human detection systems.
vehicle monitoring systems.
Keep track of information video data for future use.
Helpful in identifying people in the crime scenes etc..
Disadvantages of the present system?



It’s difficult to maintain heavy amount of raw video data
Human interaction.
Require higher bandwidth for transmitting the visual data.
3
Video surveillance in the context of
Computer Vision



Detection and tracking of moving objects are the important
tasks of the computer vision.
The video surveillance systems not only need to track the
moving objects but also interpret their patterns of behaviours.
This means solving the information and integration the
pattern.
Advantages



Minimizes the user interaction.
Less amount of prohibitive bandwidth.
Minimizes the cost and time.
4
Need for Traffic Monitoring



To reduce the traffic
congestion on highways
Reduce the road
accidents
Identifying suspicious
vehicles. Etc..,
5
Traffic Monitoring in Computer Vision

The quest for better traffic information, an increasing
reliance on traffic surveillance has resulted in a better
vehicle detection.
Taking some intelligent actions based on the conditions.

Traffic scene analysis in 3 categories.




A strait forward vehicle detection and
counting system .
Congestion monitoring and traffic scene
analysis.
Vehicle classification and tracking systems
which involve much more detailed scene
traffic analysis.
6
Responsibilities of reliable
Traffic Monitoring System






Adaptive to changes in the real world environments
Easy to set up
Capable of operating independently of human
operators.
Capable of intelligent decisions.
Capable of monitoring multiple cameras and
continuous operation.
Reasons for unsuccessful implementation**
7
A Traffic Monitoring System
8
Object Classification

Shape based classification.



Image blob area, blob bounding box
Classification based on above info.
Motion-based classification.



Human motion shows periodic property.
Time frequency analysis applied.
Residual flow taken under consideration.
9
Object tracking strategies (I)*

Background subtraction



Difference between the
current image and the
reference background
image in a pixel by pixel
fashion.
Sensitive to the background
changes
Wallflower principles for
effective background
maintenance.
10
Object tracking strategies (II)

Temporal differencing





Moving objects changes
intensity faster than static
ones
Uses consecutive frames
to identify the difference.
Adaptive to dynamic
scene changes
Problems in extracting all
relevant features.
Improved versions uses
three frames instead of
two
11
Object tracking strategies (III)

Optical flow



To identify characteristics of
flow vectors of moving objects
over time.
It’s used to detect
independently moving objects
in presence of camera.
Requires a specialized
hardware to implement.
Optical flow of moving objects
Meyer et al
12
Vehicle detection techniques




Model based detection
Region based detection
Active contour based detection
Feature based detection
13
Vehicle detection technique (I)

Model based Tracking
 The emphasis is on recovering
trajectories and models with high
accuracy for a small number of
vehicles.
 The most serious weakness of this
approach is the reliance on detailed
geometric object models.
Disadvantage
 It is unrealistic to expect detailed
models for all vehicles that could be
found on the roadway
14
Vehicle detection technique (II)

Region based tracking


It detects each vehicle blob
using a cross correlation
function.
Vehicle detection based on back
ground subtraction.
Disadvantage
 Difficult to detect the vehicles
under congested traffic,
because vehicles partly occlude
with one another
Potential segmentation problem
15
Vehicle detection technique (III)

Active contour based
detection
 Tracking is based on active
contour models, or snakes.
 Representing object in
bounding contour and keep
updating it dynamically.
 It reduced computational
complexity compared to the
region based detection.
Disadvantage:
 The inability to segment
vehicles that are partially
occluded remains a problem.
Bounding counters
16
Vehicle detection technique (IV)

Feature based detection


Tracks sub-features such as
distinguishable points or lines
on the object
Effectiveness improved by the
addition of common motion
constraint.
Features are grouped together based on
common motion, avoiding segmentation
problem due to occlusion
17
A typical vehicle tracking procedure
18
Wallflower Principles & Practice of
Background Maintenance.
•Moved objects
•Foreground capture
•Time of day
•Stopped car
•Light switch
•Moving car
•Waving trees
•Shadows
•camouflage
•Bootstrapping
19
Wallflower: Three levels of abstraction

Pixel level




Region level



Maintains models of back ground of each individual pixel.
Processing makes the preliminary classification between
foreground and background
Dynamic to scene changes.
Emphasis is on interrelationship between the pixels
Helps to refine raw classification at pixel level
Frame level

It watches for the sudden changes in the large parts of the
image and swaps in alternative background models.
20
A real time traffic monitoring system
Feature based tracking
algorithm
•Camera calibration
•Feature detection
•Vehicle tracking
•Feature grouping
Benjamin Coifman, Jitendra Malik, David Beymer
21
Offline camera definition



Line correspondences for a projective mapping.
A detection region near the image bottom and an
exit region at the image top
And multiple fiducial points for camera calibration
Based on the above information the system computes the
homography between the image coordinates(x,y) and the
world coordinates(X,Y)
22
On-line tracking and grouping

Detector



Tracker



Detecting corners at the bottom of image,
where brightness varies in more than one
direction.
Detection operationalzed by the points in
the image I
Uses kalman filters to predict the velocity
in the next image.
Normalized correlation is used to search
the small region of image.
Group



Grouper uses common motion constraint.
Once all the corner features are identified
they are grouped together.
Monitoring the distance between the
point d(t)=P1(t)-p2(t)
23
2
Sample feature tracks from the tracker
1
3
Sample corner features identified by the tracker
Sample feature groups from the tracker
24
Conclusion & Future Work

The real time traffic surveillance system is
still under research due to the background
maintenance problem and occlusion.

Better Background maintenance
Solving occlusion problem

25
References:





A Survey on visual surveillance of object motion and behaviour
– HU et al
Transportation research part-c/ A real time computer vision system for
Traffic monitoring and vehicle tracking – B.coifman, J.Malik etc..
Steps towards cognitive vision system – H.Nagel, IAKS Karlsruhe.
VSAM project – Carneigh Mellon University
Wallflower Principles and practices – Microsoft Research group.
26