detection and classification of vehicles from a video using time-spatial image nafi ur rashid,...
TRANSCRIPT
DETECTION AND CLASSIFICATION OF VEHICLES FROM A VIDEO USING TIME-
SPATIAL IMAGE
NAFI UR RASHID, NILUTHPOL CHOWDHURY, BHADHAN ROY JOYS. M. MAHBUBUR RAHMAN
Department of Electrical and Electronic EngineeringBangladesh University of Engineering And Technology
Dhaka – 1000, Bangladesh
ICECE 2010
WHY VEHICLE DETECTION AND CLASSIFICATION SYSTEM (VDCS)?
•Traffic flow parameter extraction
•Intelligent transportation system
•Automated traffic control
• Automated vehicle counting
• Automated checking of toll collection
• Toll booth – Bridges, Avenues• Parking lot – Hospital, Shopping Mall
• Detection of traffic violation
• Speed monitoring• Lane monitoring
ICECE 2010
COMMON TECHNIQUES
Mechanical techniques -
• Induction Loop Sensor
• Pneumatic Road Tube
• Weight-in-motion Sensor
• Piezoelectric Cable Sensor
ICECE 2010
LIMITATIONS
ICECE 2010
• High space requirement
• High installment & maintenance cost
• Setup & repair process time consuming
• Calibration
• Mechanical Error
• Hardware Based
ICECE 2010
SMART APPROACH - VIDEO PROCESSING
• Nonintrusive method.
• Less installation and maintenance cost.
• No disruption of traffic for installation and repair.
• Remote traffic surveillance
• Efficient classification of vehicles
• Software based
• Features & parameters are adaptive • Advanced DSP algorithms could be applied
EXISTING VIDEO-BASED DETECTION SYSTEM
Motion-based systems
• Optical Flow
• Gaussian Model
• Background Subtraction
ICECE 2010
Problems of existing systems:
• Heavy computational load• Highly sensitive to jittering & pixel intensity• Less suitable for real-time implementation
PROPOSED METHOD
VIRTUAL DETECTION LINE BASED METHOD
• Time Spatial Image (TSI) Generation• contains both temporal and spatial information• Vehicular width can be approximated
• Ensures faster extraction of Key Vehicular Frame (KVF)
• Tracking independent• Only one frame per classification• Simple yet efficient• Low computational load
ICECE 2010
VIRTUAL DETECTION LINE
Virtual Detection Line
ICECE 2010
A strip of pixel perpendicular to the direction of vehicle travelling
Back
Frame 1Frame 2
TSI GENERATION
ICECE 2010
Frame 12
TIME SPATIAL IMAGE (TSI)
ICECE 2010
Frame 692
EDGE DETECTOR
EDGE DETECTION
ICECE 2010
MORPHOLOGICAL OPERATIONS
EDGE DETECTOR
MORPHOLOGICAL OP.
ICECE 2010
ICECE 2010
Bounding Box
Center of Bounding Box
TSI PROCESSING
TSI Vehicular Blob (TVB) Width
Source video
ICECE 2010
TSI PROCESSING
Center of Bounding Box
KEY VEHICULAR FRAME
A time frame on which the midpoint of the vehicle is approximately on the VDL
• Only KVF requires further processing• No background processing required
Back
Car 1 Leguna BikeCar 2
ICECE 2010
ICECE 2010
SEGMENTATION
KVF TSI
ICECE 2010
MORPHOLOGICAL OP.
Canny Edge Detection Blob = ((Im Obj)⊕ ΘObj)Obj = 5x5 rectangle
Filling ‘holes’
FEATURE EXTRACTION
ICECE 2010
Shape-based feature Extracted from vehicle blob of TSI & KVF
Feature Selection Criteria: Distinctiveness Computational efficiency Sensitivity to environment Non-Redundancy
FEATURES
Selected Shape-Based Features:
TVB Width
Length-Width Ratio
Major Axis-Minor Axis Ratio
Area
Compactness
Solidity
.
ICECE 2010
TVB Width: Vertical length of the segmented region of TSI Vehicle Blob
FEATURES
ICECE 2010
Length-Width Ratio :
.
ICECE 2010
FEATURES
Width TVBframe selected of region segmented of Length
Major Axis-Minor Axis Ratio:
• This ellipse has the same normalized second central moments as the segmented region.
.
ICECE 2010
FEATURES
Ellipse Fitted Best of Axis-MinorEllipse Fitted Best of Axis-Major
Area:
• Number of white pixels in the segmented region.
.
ICECE 2010
FEATURES
Xmax
Xmini
Ymax
Yminj
j)Blob(i,
Compactness:
• Determines how compact(circular) a shape is.
.
ICECE 2010
FEATURES
2PerimeterArea
Solidity:
• Convex Area is the area of smallest polygon that contain the region
.
ICECE 2010
FEATURES
AreaConvex Area
ICECE 2010
FEATURE VECTOR TABLE
Width Solidity Area HWRCompact
nessAxis
Ratio
12.00 0.80 265.00 3.08 0.61 3.9916.00 0.86 335.00 2.13 0.59 2.7820.50 0.86 716.92 2.34 0.71 2.2422.42 0.88 815.38 2.34 0.67 2.3526.00 0.95 920.00 1.69 0.80 1.8830.00 0.94 975.00 1.60 0.82 2.1142.68 0.97 1714.87 1.20 0.92 1.4338.37 0.95 1471.46 1.26 0.96 1.4338.36 0.94 1654.09 1.56 0.74 1.8841.98 0.92 1985.12 1.56 0.67 1.95
Bike
Rickshaw
Auto-Rickshaw
Car
Microbus
CLASSIFICATION
K- Nearest Neighborhood (KNN) LinearWeighted Distance MeasurementMajority Voting
Why KNN?
Sufficiently low computational complexityStandard & optimumSignificantly good classification performance
.
ICECE 2010
CLASSIFICATION
Steps of obtaining Training Data Set:Feature vectors were obtained from handpicked vehiclesObtained feature vectors were partitioned with Fuzzy C-
Means Clustering algorithm
Why FCM?
Reduction of memory requirement Speeding up of searching time
Majority voting among the training data set determines vehicle class
ICECE 2010
ICECE 2010
EXPERIMENTAL SETUP
• Camera Elevation: 21 feet
• Camera Angle: 45 degrees
• Frame Rate: 25 fps
• Resolution: 144x176 pixels
• Color Profile: Monochrome
• Weather Condition: Sunny, Cloudy, Normal
• System Specification: Intel Pentium D 2.66 GHz, 1GB DDR2 RAM
ICECE 2010
BLOCK DIAGRAM
Video-Input Extracted
FramesKVF Extraction
Feature Extraction
Blob Detection
Object Detection
KNN
Class Type
Training Dataset
TSI Generation
VDL
ICECE 2010
EXPERIMENTAL DATA
Method [1]:
ICPR 2002,IICETC 2009
Method [10]:
In. J. Intel. Eng. Sys. 2009
Total Frames: 3082 (Sequence 1)Method [1]: 35.4s Proposed Method: 10.3s
FUTURE WORK
Introduction of texture based & motion invariant features to reduce classification errors
Multiple VDL
• Speed Calculation• Improved detection & classification• Occlusion minimization
ICECE 2010
CONCLUSION
Significant improvement in terms computational load
Efficient designing of intelligent transportation system
Significantly low misclassification & misdetection rate than that of traditional methods
Practically implementable in many important sectors
ICECE 2010
THANK YOU………