Abstract: Road detection is an important part of the intelligent transportation vision system, according to the current problems of real-time, accuracy, robustness for unstructured road detection in complex environment, a new method is proposed for road detection. The means compresses data information of image through Gaussian pyramid down-sampling process and adopts bilateral filtering to suppress noise, then extracts edges of the filtered images based on modulus maximum of wavelet transform, uses threshold method to remove non-road edge points. A new K-means clustering algorithm is proposed which is based on slope and intercept, and it realizes road equation fitting. Experimental results show that this method can realize unstructured road detection more accurately in complicated road scene and improve real-time than traditional methods.

Key words: road detection, Gaussian pyramid, bilateral filtering, wavelet transform, modulus maximum, K-means clustering

摘要： 道路检测是智能交通视觉系统的一个重要组成部分，为提高复杂环境下非结构化道路检测的实时性、准确性和鲁棒性，提出一种新的道路检测方法。该方法利用高斯金字塔对图像进行降采样，压缩图像数据信息，对图像进行双边滤波，抑制噪声，采用基于小波变换求模极大值的方法对滤波后的图像提取边缘，通过阈值法去除非道路边缘点，给出基于斜率和截距的K-means聚类算法，实现道路方程拟合。实验结果表明，与传统最小二乘法相比，该方法能在道路场景较为复杂的情况下更准确地实现非结构化道路检测，并提高实时性。

关键词: 道路检测, 高斯金字塔, 双边滤波, 小波变换, 模极大值, K-means聚类

CLC Number: 

• TP391