Vehicle-mounted surround vision algorithm based on heterogeneous architecture
- PDF / 4,854,835 Bytes
- 24 Pages / 439.37 x 666.142 pts Page_size
- 52 Downloads / 151 Views
Vehicle-mounted surround vision algorithm based on heterogeneous architecture Tong Liu 1 & Jindong Zhang 1,2,3 & Kunpeng Zhang 1 & Jiabin Xu 1 & Donghui Wang 4 & Xue Wang 4 Received: 10 July 2019 / Revised: 24 April 2020 / Accepted: 11 June 2020 # Springer Science+Business Media, LLC, part of Springer Nature 2020
Abstract
In order to take advantage of the powerful advantages of heterogeneous devices and improve the robustness of the vehicle-mounted surround vision algorithm(VSVA), several key technologies of VSVA are improved in the paper. Firstly, computationally intensive tasks are calculated by heterogeneous Graphics Processing Unit(GPU), at the same time, so as to adapt to the VSVA, the memory model and computing model of GPU are optimized. Then a perspective transformation algorithm based on geometric constraints is proposed to improve the quality of the transformed image. Finally, an image alignment and fusion algorithm based on a calibration board is proposed, which reduces the complexity of the algorithm while ensuring the robustness of the image fusion algorithm. The paper compares the proposed algorithm with the traditional algorithm, the test results show that the proposed algorithm has good robustness and the overall performance of the VSVA is improved to 95.39%, the proposed algorithm can be widely used. Keywords VSVA . GPU . Computationally intensive task . Robustness
1 Introduction In Advanced Driver Assistance System (ADAS), vehicle vision has attracted more and more attention from researchers [2, 18]. If we can better use visual sensors to obtain scene information during the implementation of ADAS, it will speed up the intelligent process of ADAS and reduce the cost of ADAS rapidly. In order to make better use of on-board vision to provide users with a more comfortable and safe experience, many companies and researchers have made great efforts in this regard. Among them, the ADAS-based vehicle-mounted surround vision algorithm (VSVA) has attracted widespread attention because it can eliminate all blind spots near the vehicle body and provide all the scenes around the vehicle [3, 12]. * Jindong Zhang [email protected] Extended author information available on the last page of the article
Multimedia Tools and Applications
As the most basic part of ADAS vehicle vision [17], VSVA acquires initial scene information through four fish-eye cameras around the vehicle body, and then processs and transforms it to form a 360 degree blind-free panoramic image [6, 13, 20]. After calibrating the fish-eye camera, Wang et al. [15] used an image mosaic method based on LevenbergMarquardt algorithm form the onboard surround view, and developed a low-cost bird’s eye view vision assistance system. In the process of image mosaic, Gao et al. [5] used threedimensional ship model and texture mapping form a 3-D surround view, and realized 3-D output display. Hsu et al. [8] proposed a new framework of fish eye correction algorithm and panoramic mosaic algorithm, which can be applied to VSVA. Lin et al. [11] propos
Data Loading...
