An End-to-End Solution to Autonomous Driving Based on Xilinx FPGA
Authors:
- 1 December 2019
- conference paper
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- p. 427-430
Abstract
Nowadays, the autonomous driving topic is very hot, many people are trying to provide a solution to this problem. This time we build our own auto-driving car based on Xilinx Pynq-Z2, it provides an end-to-end solution which inputs images from camera and outputs control instructions directly. The platform also uses the power of Deep learning Processing Unit(DPU) to accelerate the inference process and provides a simulator for training and testing in virtual environment. If the car meets some situations which cannot be handled by AI model, it's easy to add extra traditional computer vision functions to our control system. So our platform can help people who want to try autonomous driving build their own model and test it efficiently. We hope that our platform can be easy to use and extend.Keywords
This publication has 10 references indexed in Scilit:
- Development of an Autonomous Driving Robot Car Using FPGA2018
- In-Datacenter Performance Analysis of a Tensor Processing Unit2017
- Accelerating Binarized Neural Networks: Comparison of FPGA, CPU, GPU, and ASIC2016
- TABLA: A unified template-based framework for accelerating statistical machine learning2016
- An approach of lane detection based on Inverse Perspective Mapping2014
- Object Detection using Circular Hough TransformAmerican Journal of Applied Sciences, 2005
- A 3D-Polar Coordinate Colour Representation Well Adapted to Image Analysis2003
- A Computational Approach to Edge DetectionIEEE Transactions on Pattern Analysis and Machine Intelligence, 1986
- Extracting Straight LinesIEEE Transactions on Pattern Analysis and Machine Intelligence, 1986
- K-Means-Type Algorithms: A Generalized Convergence Theorem and Characterization of Local OptimalityIEEE Transactions on Pattern Analysis and Machine Intelligence, 1984