A complete prototype system for measuring vehicle lateral position has been set up during the course of this master’s thesis project. In the development of the software, images acquired from a back-ward looking video camera mounted on the roof of the vehicle were used.The problem of using computer vision to measure lateral position can be divided into road marking detection and lateral position extraction. Since the strongest characteristic of a road marking image are the edges of the road markings, the road marking detection step is based on edge detection. For the detection of the straight edge lines a Hough based method was chosen. Due to peak spreading in Hough space, the difficulty of detecting the correct peak in Hough space was encountered. A flexible Hough peak detection algorithm was developed based on an adaptive window that takes peak spreading into account…
Contents
1 INTRODUCTION
1.1 BACKGROUND AND MOTIVATION
1.2 PROBLEM DEFINITION
1.3 SYSTEM SPECIFICATIONS AND REQUIREMENTS
1.4 DELIMITATIONS
1.5 METHOD
1.6 SYSTEM OVERVIEW
1.7 OUTLINE OF THE THESIS
2 BACKGROUND AND PREVIOUS WORK
2.1 LATERAL POSITIONING
2.1.1 Time-to-Line Crossing (TLC) and other Applications6
2.2 VISION-BASED APPROACH
2.3 IMAGE ACQUISITION
2.3.1 Camera
2.3.2 Camera Mounting
2.3.3 Field of View
2.4 ROAD MARKING DETECTION
2.4.1 General Problems
2.4.2 Typical Assumptions
2.5 ALGORITHMS FOR ROAD MARKING DETECTION
2.5.1 The Gradient and Edge Detection
2.5.2 The Hough Transform
2.5.3 Colour Image Processing
2.5.4 Temporal Correlation in Image Sequences
2.6 GEOMETRICAL MODEL
2.7 CAMERA CALIBRATION
2.7.1 Look-up Table
2.7.2 Plane Projective Mapping
2.7.3 Angle Extraction
3 SYSTEM DESIGN APPROACH
3.1 ROAD ATTRIBUTES
3.1.1 Road Marking Characteristics
3.1.2 Road Characteristics
3.2 PROTOTYPE SYSTEM SET-UP
3.2.1 System Overview
3.2.2 The Image Acquisition System
3.2.3 Camera Mounting and Field of View
3.2.4 Road Marking Detection/Functionality
3.2.5 Camera Calibration
3.2.6 Lateral Position and Yaw Angle Extraction
3.2.7 Algorithm
3.3 POTENTIAL PROBLEMS
3.3.1 Dynamic errors
3.3.2 Curves and Perspective effect in image
3.3.3 Intermittent road markings
4 ROAD MARKING DETECTION
4.1 PRE-PROCESSING
4.2 ROAD MARKING TRACKING
4.3 HOUGH TRANSFORMATION
4.3.1 Characteristic Point Detection
4.3.2 Transform Mapping
4.3.3 Peak Detection Methods
4.3.4 Back-projection
4.3.5 Verification of Road Marking Candidates
4.3.6 Radon Transform
5 SYSTEM CALIBRATION AND DATA EXTRACTION
5.1 CAMERA SYSTEM CALIBRATION
5.1.1 Calibration Measurements
5.2 CAMERA MODEL
5.3 MAPPING TO REAL-WORLD COORDINATES
5.4 YAW ANGLE EXTRACTION
5.5 LATERAL POSITION EXTRACTION
6 EXPERIMENTS AND RESULTS
6.1 DATA EXTRACTION VALIDATION
6.2 ROAD MARKING TRACKING EXPERIMENTS
6.2.1 Rate of valid data
6.2.2 Continuous road markings
6.2.3 Other types of road markings
6.2.4 Shadows
6.2.5 Lane change manoeuvre
7 DISCUSSION AND CONCLUSIONS
7.1 DISCUSSION
7.2 PROBLEM ANALYSIS
7.2.1 Sources of Error
7.3 CONCLUSIONS
7.4 FUTURE IMPROVEMENTS
8 REFERENCES
APPENDICES
Author: Ågren, Elisabeth
Source: Linköping University
Download URL 2: Visit Now