DEPARTMENT OF INFORMATION TECHNOLOGY
Lane Detection in Videos Using Image
Processing Techniques
(Font- Times New Roman, 18 pts)
Under The Esteemed Guidance of
Dr. KUPPANI SATHISH, M. Tech., PhD,
Professor
Submitted By: -
Name: Roll No:
NALLAPATI HARSHA VARDHAN 21NE1A1286
RAPATHATI OBULESU 21NE1A12B6
MITTA MADHU BABU 21NE1A1279
SHAIK NAGUL MEERA 21NE1A12B7
ABSTRACT (Font- Times New Roman, 12 pts, BOLD)
The Lane Detection system is a computer vision application for identifying and tracking Lane lines in
video footage. It processes real-time frames to enhance road safety and assist autonomous vehicle
navigation. The system effectively highlights lane boundaries by leveraging color filtering, edge
detection, and Hough transform, providing a foundation for advanced driver-assistance systems
(ADAS) and traffic monitoring.
EXISTING SYSTEM:
Existing lane detection systems often rely on basic techniques like static edge detection or manual
identification of lane boundaries, which are susceptible to inaccuracies under challenging conditions
such as low lighting, shadows, or varying road types. Many existing solutions fail to adapt to dynamic
environments, resulting in unstable lane tracking and a lack of real-time processing capabilities.
Limitations of the Existing System:
• Inconsistent detection under poor lighting or weather conditions.
• Inefficient for real-time video processing.
• Lack of adaptive smoothing for lane line stability.
PROPOSED SYSTEM:
The proposed system addresses the limitations of existing lane detection approaches by implementing a
robust and efficient algorithm that processes video frames in real time. Key enhancements include
adaptive lane line smoothing, dynamic region-of-interest selection, and noise reduction to ensure
consistent detection under various conditions.
MODULE DESCRIPTION:
Number of Modules
After careful analysis, the system has been identified to have the following modules:
1. Preprocessing Module
2. Edge Detection Module
3. Region of Interest Module
4. Line Detection and Classification Modul
5. Smoothing and Visualization Module
Modules:
1. Preprocessing Module:
o Converts frames to grayscale and HSV color spaces.
o Applies color thresholding to isolate lane-like regions.
2. Edge Detection Module:
o Reduces noise using Gaussian blur.
o Extracts edges using the Canny edge detection algorithm.
3. Region of Interest Module:
o Defines a polygonal region to focus on road lanes while ignoring irrelevant
areas.
4. Line Detection and Classification Module:
o Detects lines using the Hough Transform.
o Classifies lines into left and right lanes based on slope and position.
5. Smoothing and Visualization Module:
o Averages detected lanes across frames for smoother visualization.
o Overlays detected lanes on the original frame.
SOFTWARE REQUIREMENTS:
Programming Language: Python 3.x.
Libraries: >OpenCV: For Image Processing
>NumPy: For Numerical Computations
>MoviePy: For Video Processing
Operating System: Windows/Linux
IDE/Editor: VS code, IDLE, or Jupyter Notebook
HARDWARE REQUIREMENTS:
Processor: Intel i5 or above.
RAM: 8 GB or higher.
Storage: 500 GB HDD or SSD.
Graphics: Integrated GPU or better for faster processing.
Camera: HD camera for video input (if using live feed).
References
• P. V. P. Padmaraja, R. Rohith, and S. Chittesh, "Lane Detection using Image Processing for
Autonomous Vehicles," in 2023 2nd International Conference on Advancements in Electrical,
Electronics, Communication, Computing and Automation (ICAECA), 2023.
• M. Philip, M. A. Kurian, R. Joseph, R. Sruthi, and S. Thomas, "A Computer Vision Approach
for Lane Detection and Tracking," in 2021 2nd International Conference on Advances in
Computing, Communication, Embedded and Secure Systems (ACCESS), 2021.