UFL-HS-RoadDetection

Code for our paper

Robust Road Detection from a Single Image.
Junkang Zhang, Siyu Xia, Kaiyue Lu, Hong Pan, and A. K. Qin.  
23rd International Conference on Pattern Recognition (ICPR), 2016.

Introduction

This code has been tested using Matlab 2015b in Ubuntu 14.04 64-bit. It should also work on other platforms, since it's pure Matlab code.

Usage

For example, to get the result on After-Rain dataset

0. Download the dataset from http://www.josemalvarez.net/ and extract it to someplace.

1. Open main.m, change the route variable (in line 7) to the location of your dataset (where the extracted After-Rain folder lies).

2. (Optional) When with a multi-core cpu and sufficient memory, one can use parfor (in line #60) to speed up the testing process.

3. Run main.m. The P, R, F values will be printed in the command window. The detection results are stored in the predCell variable.