A Novel Monocular Camera Obstacle Perception Algorithm for Real-Time Assist in Autonomous Vehicles

Gadagkar, AV and Umanand, L (2020) A Novel Monocular Camera Obstacle Perception Algorithm for Real-Time Assist in Autonomous Vehicles. In: 6th IEEE International Symposium on Smart Electronic Systems, iSES 2020, 14-16 Dec 2020, Chennai, pp. 196-201.

PDF ises_2020.pdf - Published Version Restricted to Registered users only Download (312kB) | Request a copy

Abstract

This paper aims at reducing the processing required for manoeuvring of autonomous vehicle by a novel image processing algorithm for obstacle detection and localization. A monocular camera has less data overhead and bypasses the issues of synchronization and stereo calibration. However, estimating 3D information using a single image from a monocular camera is non-Trivial primarily due to ambiguity in the scale of the scene arising from limited information of the surroundings. This paper proposes a novel 'Calibration Lane a low-latency real-Time algorithm for lane assist, obstacle detection and avoidance using a single measurement in world coordinates. It allows the perception pipeline to be extremely light-weight and requiring a single monocular camera. The algorithm is validated and verified experimentally. © 2020 IEEE.

Item Type:	Conference Paper
Publication:	Proceedings - 2020 6th IEEE International Symposium on Smart Electronic Systems, iSES 2020
Publisher:	Institute of Electrical and Electronics Engineers Inc.
Additional Information:	The copyright for this article belongs to Institute of Electrical and Electronics Engineers Inc.
Keywords:	Automobile electronic equipment; Calibration; Cameras; Obstacle detectors; Stereo image processing, Image processing algorithm; Limited information; Monocular cameras; Obstacle detection; Obstacle perception; Real time algorithms; Stereo calibration; World coordinates, Autonomous vehicles
Department/Centre:	Division of Electrical Sciences > Electronic Systems Engineering (Formerly Centre for Electronic Design & Technology)
Date Deposited:	18 Aug 2021 07:54
Last Modified:	18 Aug 2021 07:54
URI:	http://eprints.iisc.ac.in/id/eprint/69281

Actions (login required)

View Item