Modeling Design and Analysis of Intelligent Traffic Control System Based on Statistical Image Processing Techniques
Download as PPTX, PDF4 likes1,362 views
The document presents a model for an intelligent traffic control system utilizing statistical image processing techniques for monitoring and managing intercity traffic. It outlines the design and functioning of vehicle detection, counting, and classification systems, culminating in data display and traffic signal control. Experimental results indicate the system's effectiveness in real-time traffic management, though it notes areas for further improvement.
Modeling Design and Analysis of Intelligent Traffic Control System Based on Statistical Image Processing Techniques
- 1. Modeling, Design and Analysis of Intelligent Traffic Control System Based on Integrated Statistical Image Processing Techniques Yasar Abbas Ur Rehman Department of Electrical Engineering City University Peshawar FAST, NUCES Peshawar Campus
- 2. Outline • Introduction • Problem Statement • Proposed Solution • Results • Conclusion 2
- 3. Introduction • Monitoring and control of intercity traffic ▫ Not a trivial problem ▫ Careful planning ▫ Increase in road infrastructure? ▫ Availability of Technological Assets • Camera controlled monitoring and control ▫ Vehicle flow ▫ Speed calculation ▫ Automatic Number Plate Recognition (ANPR) ▫ Crash detection 3
- 4. Problem Statement • Detection Problem ▫ Presence and absence of vehicle ▫ No background information • Design of Autonomous System for Traffic ▫ Detection ▫ Classification ▫ Display 4
- 5. Proposed Solution • Detection Problem ▫ Background Modeling 𝐵𝑖 = 𝛼 × 𝐵𝑜 + 1 − 𝛼 𝐼𝑖, 0 ≤ 𝛼 ≤ 1 (1) 5
- 6. Vehicle Detection 𝑃𝑖 𝑘 𝐿 = 𝑛 𝐿 𝑁 (2) 𝑃𝑖+1 / 𝑘 𝐿 = 𝑛 𝐿 / 𝑁 (3) ∆𝑃𝐿 = 𝑃𝑖+1 / 𝑘 𝐿 − 𝑃𝑖 𝑘 𝐿 (4) 𝑇 = 𝑁 𝐿=0 255 ∆𝑃𝐿 (5) To simplify calculations, we make a use of histogram and relate it with the above equations: ℎ𝑖 𝑘 𝐿 = 𝑛 𝐿 (6) ℎ𝑖 𝑘 𝐿 = 𝑁 × 𝑃𝑖 𝑘 𝐿 (7) 6
- 7. Cont. 𝑃𝑖+1 / 𝑘 𝐿 − 𝑃𝑖 𝑘 𝐿 = | ℎ 𝑖 𝑘 𝐿 − ℎ 𝑖 / 𝑘 𝐿 | 𝑁 (8) ∆𝑃𝐿 = | ℎ 𝑖 𝑘 𝐿 − ℎ 𝑖 / 𝑘 𝐿 | 𝑁 (9) 𝑇 = 1 𝑁 𝐿=0 255 | ℎ𝑖 𝑘 𝐿 − ℎ𝑖+1 / 𝑘 𝐿 | (10) 𝑉𝑖 = 1, 𝑇 > 𝑡 0, 𝑒𝑙𝑠𝑒 (11) 7
- 8. System Design • Vehicle Detection System (VDS) • Vehicle Counting and Classification System (VCCS) • Traffic Signals Control System (TSCS) • Data Display System (DDS) 8
- 10. VDS • Probability Based Vehicle Detection (PBVD) Algorithm ▫ Frame acquisition ▫ Vehicle appearance 𝑇 = 1 𝑁 𝐿=0 255 | ℎ𝑖 𝑘 𝐿 − ℎ𝑖+1 / 𝑘 𝐿 | (10) 𝑉𝑖 = 1, 𝑇 > 𝑡 0, 𝑒𝑙𝑠𝑒 (11) ▫ Vehicle extraction ▫ Maximum area calculation 10
- 11. VCCS • Vehicle labeling • Classification of vehicles ▫ Small ▫ Medium ▫ Large • Send vehicle statistics to ▫ TSCS ▫ DDS 11
- 12. TSCS • Vehicle comparison for both lanes • If same number of vehicles ▫ Assign equal timing to both lanes 12
- 13. DDS • Display total number of vehicles • Number of pixels each vehicle contain • Vehicle category • Green signal for road 13
- 14. Experimental Results • Background update parameter ▫ α = 1 × 10-3 • Threshold for vehicle appearance ▫ t = 300 14
- 15. Cont.. Prototype of Proposed System 15
- 16. Cont. Data Display System (DDS) 16
- 17. Cont. a) Image of empty road b) image of the road containing car c) image enhancement after median filtering d) image after frame subtraction 17
- 18. Cont. a) Image after edge detection b) Binary dilation c) Binary Hole Filling d) Binary Erosion 18
- 19. Cont. a) Area Calculation b) Final Result (Detected Objects) 19
- 22. Cont. Edge detection and binary dilation 22
- 23. Cont. Resultant image after binary hole filling, erosion and area calculation 23
- 25. Cont.. 0 50 100 150 200 250 0 5 10 15 20 25 30 Pixel intensities Absolutedifferene Spred after absolute histogram subtraction Spread after absolute histogram subtraction 25
- 26. Conclusion • Traffic control system for ▫ Monitoring ▫ Control of intercity traffic • Prototype testing ▫ Accurate results • Real time testing ▫ Satisfactory results ▫ Morphological operators need control ▫ Need of restoration techniques 26
- 27. Question 27