UAV-assisted Deep Vision Vehicles Management System for Karachi, Pakistan

다국어 초록 (Multilingual Abstract)

These days, the advancement of Information and Communication Technology (ICT) has affected our daily lives significantly, including governance. Urban public infrastructure facilities including Transportation Systems, Water and Sewer networks and energy supplies are getting intelligent using open-source data-driven platforms in the modern world. Unfortunately due to lack of resources and poor governance in third world countries, the local circumstances of the cities like Karachi have not been extensively explored by researchers despite of growing population. This thesis mainly discusses the viability of using computer vision and unmanned aerial vehicles in object detection for congested roads of Karachi. The famous CNN algorithms like YOLOv5 and Cascade classifier are modified and used for vehicles running on the overcrowded road for accurate recognition, counting and speed anomalies detection. Similarly, their License Plates (LP) have been recognizes using Optical Character Recognition specifically considering the local parameters. A comparison of accuracy, robustness in different surroundings and detection speed with Cascade Classifier and Tesseract OCR has been performed. Custom model training on YOLOv5 has been performed and an ‘Accuracy Enhancement’ technique integrating real-time drone stream data with YOLOv5 and easy OCR using a mean averaging algorithm has been proposed to get better accuracy in license plate detection.

목차 (Table of Contents)

• LIST OF FIGURES iii
• LIST OF TABLES v
• LIST OF ACRONYMS vi
• ABSTRACT ix
• CHAPTER 1 INTRODUCTION 1
• 1.1 Overview 1
• 1.2 Motivation 2
• 1.3 Research Objective 3
• 1.3.1 Broad Goal 3
• 1.3.2 Specific Research Question 3
• 1.4 Problem Description 3
• 1.4.1 Data-Driven Framework for Urban Facilities 5
• 1.4.2 Image Processing and Smart Cities 5
• 1.4.3 Unmanned Aerial Vehicles and Smart Cities 6
• CHAPTER 2 LITERATURE REVIEW 7
• 2.1 Unmanned Aerial Vehicles in Computer vision Object Detection 7
• 2.2 Data-Driven Approach 11
• 2.2.1 Artificial Intelligence 11
• 2.2.2 Big Data 13
• 2.2.3 Information and Communication Technology 14
• 2.3 Data-Driven Urban Mobility 15
• 2.4 Global Data-Driven Mobility Experience 17
• 2.5 Seoul City Data Governance Model 20
• 2.5.1 Historical Evolution of data in Seoul Governance 20
• 2.5.2 Key institutions 21
• 2.5.3 Seoul Transport Operation & Information Service (TOPIS, 2021) 21
• 2.6 Computer Vision in ITS 25
• 2.6.1 Object and Characters detection in Deep Learning 25
• 2.6.3 You Only Look Once 27
• 2.6.4 Easy OCR 27
• 2.6.5 Py-Tesseract 28
• CHAPTER 3 RESEARCH METHODOLOGY 29
• 3.1 Drone Technology in Karachi Traffic Surveillance 29
• 3.2 Data Acquisition 30
• 3.3 Testing System Specification 31
• 3.4 Research Pipeline 32
• 3.5 YOLOv5 Architect 32
• 3.6 Custom Training of YOLOv5 on vehicles Data 35
• 3.7 Contour Detection 37
• 3.8 Easy Optical Character Recognition 38
• 3.9 Accuracy Enhancement using the mean averaging algorithm 39
• 3.10 Cascade Classifier 42
• 3.11 Python Tesseract 43
• 3.12 Detections Using Cascade Classifier and Python Tesseract 43
• CHAPTER 4 CONCLUSION 45
• REFERENCES 47