Vision-Based System for Road Lane Detection and Lane Type Classification

Sie sind hier: Homepage > Suche

Vision-Based System for Road Lane Detection and Lane Type Classification

Madake, Jyoti / Gupta, Dhavanit / Bhatlawande, Shripad / Shilaskar, Swati

This study proposes a method for the detection and classification of road lane boundaries. In an autonomous vehicle, the scene understanding of the road in front of the vehicle is most essential. For safe driving, only detection of lanes is not sufficient, the vehicle position with respect to the road boundary geometry, road borders are also essential. The classification of road lane type is very important for the safety of passengers and for the management of traffic. The proposed system can decrease the number of accidents and increase the safety of passengers. This system is implemented using two modules. The first module detects the road lanes and color them, and the second module classifies the lane into its type. Hough transform is used to recognize curved and straight lane. After detection, the lanes are color-coded to improve identification. For lane classification, a CNN model is trained which gives a very good performance. This model is created to classify three different types of markings such as white dashed, solid white, and double yellow with 97.5% accuracy.

Zugriff

Zugriff prüfen

Verfügbarkeit in meiner Bibliothek prüfen

Bestellung bei Subito €

Exportieren, teilen und zitieren

Dokumentinformationen

Titel :

Vision-Based System for Road Lane Detection and Lane Type Classification

Weitere Titelangaben:

Lect. Notes in Networks, Syst.

Beteiligte:

Mahapatra, Rajendra Prasad (Herausgeber:in) / Peddoju, Sateesh K. (Herausgeber:in) / Roy, Sudip (Herausgeber:in) / Parwekar, Pritee (Herausgeber:in) / Madake, Jyoti (Autor:in) / Gupta, Dhavanit (Autor:in) / Bhatlawande, Shripad (Autor:in) / Shilaskar, Swati (Autor:in)

Erschienen in:

Proceedings of International Conference on Recent Trends in Computing ; Kapitel : 38 ; 441-452

Lecture Notes in Networks and Systems ; 600

Erscheinungsdatum :

2023-03-21

Format / Umfang :

12 pages

ISBN :

978-981-19-8825-7 , 978-981-19-8824-0

ISSN :

2367-3389 , 2367-3370

DOI :