International Research Journal of Engineering and Technology (IRJET) Volume: 11 Issue: 03 | Mar 2024
www.irjet.net
e-ISSN: 2395-0056 p-ISSN: 2395-0072
DRIVING AUTOMATION USING LANE DETECTION ALGORITHM P. Mansa Devi1, Riddhi Chakrabarti 2, Vineet Mathireddi3, Alok Kumar4 and Ravi Kiran5 1 Assistant Professor, Dept. of CSE, GITAM (Deemed to be University), Visakhapatnam, Andhra Pradesh, India. 2345Student, GITAM (Deemed to be University), Visakhapatnam, Andhra Pradesh, India.
---------------------------------------------------------------------***---------------------------------------------------------------------
Abstract - The project "Driving Automation Using Lane
Detection Algorithms" aims to revolutionize the automotive industry by harnessing the power of computer vision and machine learning. With the proliferation of autonomous vehicles, lane detection plays a pivotal role in ensuring safe and reliable self-driving capabilities. Our research focuses on developing state-of-the-art lane detection algorithms that can accurately identify and track road lanes under diverse and challenging real-world conditions. By leveraging advanced image processing techniques and deep learning models, we seek to provide vehicles with the ability to interpret complex road environments, including highways, urban streets, and adverse weather scenarios. This abstract explores the integration of Canny Edge Detection for lane detection in the context of driving automation. Lane detection is a critical component of autonomous vehicles and advanced driver assistance systems (ADAS), enabling precise vehicle positioning within road lanes. Canny Edge Detection, a classical computer vision technique, is employed to identify lane boundaries, offering a computationally efficient and robust solution. This abstract provides an overview of the approach's potential to contribute to safer and more efficient autonomous driving systems by leveraging the strengths of Canny Edge Detection for comprehensive scene understanding and accurate lane detection. Key Words: Driving Automation, Lane detection Algorithm, Autonomous Vehicles, Canny Edge Detection, Advance Driver Assistance Systems (ADAS), Road Lane Detection
method aids in helping cars recognize where they are in relation to lane boundaries and make judgment calls to stay in their designated lanes. The captured video frames were processed using the Canny edge detection algorithm to look for edges. After that, additional processing was done on these edges to identify lane markings. By implementing this plan, we hope to develop a lane-detection system that is dependable and advances the field of autonomous driving technology.
2. LITERATURE REVIEW Vision-Based Robust Lane Detection and Tracking in Challenging Conditions: The first paper proposes a method using three key technologies to detect road lane markings under difficult conditions such as changing illumination. First, to handle different types of lane edges, edge detection is improved by the Comprehensive Intensity Threshold Range (CITR). Second, the Hough Transform and geometric constraints are used by the Two-Step Lane Verification Technique to confirm the lane characteristics. Finally, even in situations where markings are partially or completely invisible, the Novel Lane Tracking Method forecasts lane positions by using historical frames. Testing on various datasets demonstrates detection rates of 97.55% and processing times of 22.33 ms per frame, which are faster than state-of-the-art techniques. The algorithm performs well in real-world lane marking detection scenarios, especially when there is noise present.
Using computer vision and machine learning, the groundbreaking project "Driving Automation Using Lane Detection Algorithms" aims to revolutionize the field of autonomous driving. To drive safely on roads, humans depend on their ability to see and obey lane markings. Similarly, autonomous cars need to be able recognize and follow these cues with unparalleled accuracy. There are currently two accepted approaches: the model-based approach and the feature-based method. There are currently two well-established techniques for using video to perform lane recognition: the feature-based method and the model-based method. Our project's mainstay, canny edge detection, is essential for correctly extracting lane markings from video streams. When combined with other image processing and computer vision algorithms, this
Vehicle Lane Change Prediction on Highways Using Efficient Environment Representation and Deep Learning: The second paper introduces a novel lane detection method using light field (LF) technology to improve prediction accuracy and robustness in intelligent transportation systems. Deep convolutional neural networks (CNNs) are promising for lane detection, but they have difficulty generalizing to different types of road conditions. The LF-based method offers better performance and robustness by making use of additional angular information that LF captures. Experimental results demonstrate the effectiveness of the proposed method in improving lane line point prediction accuracy and robustness against adverse conditions, outperforming traditional image-based lane detection methods. All things considered, the LF-based lane detection approach holds promise for resolving issues with conventional techniques and enhancing prediction robustness and accuracy.
© 2024, IRJET
ISO 9001:2008 Certified Journal
1. INTRODUCTION
|
Impact Factor value: 8.226
|
|
Page 1477