1. ANGADI INSTITUTE OF TECHOLOGY AND
MANAGEMENT
BELGAUM
VISION BASED REAL-TIME
SYSTEM FOR TRAFFIC
ANALYSIS
GROUP MEMBERS:
1.Mahaveer 2.Suraj C.
3.Mahesh P.
2. INTRODUCTION:
• Traffic problems nowadays are increasing because of
growing number of vehicles.
• Automatic traffic monitoring and surveillance are
important for traffic usage and management.
• traffic control techniques are
1.magnetic loop 2. infra-red and radar sensors
• It is well recognized that vision-based camera system
are more versatile for traffic parameter estimation.
• Image tracking of moving vehicles can give us
quantitative description of traffic flow.
3. WHY TRAFFIC ANALYSIS
NECESSARY?
• Fundamental importance for traffic operation,
pavement design, enhance public safety and
transportation planning.
• Highway Capacity Manual requires adjustments
to heavy-vehicle volumes in capacity analysis.
4. PROPOSED SYSTEM:
• It is a prototype design
and development of
vision based real-time
vehicle counting &
classification system.
• The system will detect
vehicles through video.
• The video will then be
analyzed for vehicle
detection &
classification.
• The system contains
5. PROPOSED METHODS:
1.VEHICLE COUNTING:
1.Background estimation
2.Thresholding
3.Vehicle detection
• When vehicle enters into the ROI the
vehicle is counted by Vision system once.
• The proposed vision system is able to
count the total number of vehicles which
pass through the ROI.
• Vehicle passed in lane can be estimated.
6. 2.VEHICLE CLASSIFICATION:
• The detected vehicle is sent to blob analysis
method in which statistical features are
extracted to classify vehicles.
• The data will be saved on database for further
use or analysis.
• Thresholding is performed in order to obtain
presence/absence of information.
7. THE PROPOSED SYSTEM
CONTAINS:
• User login :User on site
• Administrator login :Database change
• Video preview : Video information
• Camera calibration: Region of interest
• Camera control :Brightness, contrast and color
8. • Process start and stop
• Process scheduler: Predefined time
• Report generation: Report in excel or pdf
format.
• Option to create Image database of detected
vehicles
• Video save option
9. PROJECT OUTCOMES:
• Count of vehicle
• Type of vehicle (bus, car)
• Time at which vehicle passed on the road.
• Image database of detected vehicles
• Report generation in .pdf/.xls
11. Future implement
• Occlusion detection.
• Light reflection should be investigated and
explored to enhance the reliability of the system.
• Speed of vehicle can be measured.
12. Future implement
• Improve the algorithm to work in night and all
weather conditions.
• Traffic jam prediction.
• License plate recognition system to be
embedded in the algorithm.