Author:
Liang Zuowei,Chen Lei,An Fengwei
Abstract
Abstract
Although surround view systems have been widely used in advanced automotive, processing speed is still a bottleneck for the embedded system. This paper presents a 3D surround-view algorithm suitable for embedded platforms. Using methods such as extracting regions of interest (ROI) and HSV color screening to improve the efficiency of image registration, there is no need to manually select feature points, which reduces the difficulty of operation. In addition, we designed a bowl-shaped three-dimensional model to be used as a projection surface. In terms of optimization, methods such as look-up table (LUT) and hardware acceleration significantly reduce computational time.
Subject
General Physics and Astronomy
Reference7 articles.
1. Talking about the principle and application of automobile advanced driver assistance system (ADAS) [J];Liu;Engineering Technology Full Text Edition,2017
2. A Low-Cost Solution for An Integrated Multisensor Lane Departure Warning System[J];Clanton;IEEE Transactions on Intelligent Transportation Systems,2009
3. Photometric alignment for surround view camera system
4. Sven Loncaric. A System for Vehicle Surround View[J];Hedi;IFAC Proceedings,2012
5. A generic camera model and calibration method for conventional, wide-angle, and fisheye lenses.[J];Juho;IEEE Transactions on Pattern Analysis and Machine Intelligence,2006