基于Kmeans的道路背景检测算法研究

Research on Road Background Detection Algorithm Based on Kmeans

笔者针对高速公路卡口抓拍的车辆图像,研究了一种利用Kmeans聚类算法进行道路背景检测的方法。使用Kmeans算法对灰度图像的灰度值进行三类聚类,并计算3个类别像素所占的比例,以占比最大的类别作为道路背景类别。实验表明,设计算法的背景检测成功率更高,能够为高速公路卡口抓拍的车辆图像进行道路背景检测提供新思路。

Aiming at the vehicle images captured at the highway bayonet,this paper studies a method of road background detection using kmeans clustering algorithm.The kmeans algorithm is used to cluster the gray values of the gray image,and the proportion of the three categories of pixels is calculated,and the category with the largest proportion is taken as the road background category.Experimental results show that the design algorithm has a higher success rate of background detection,which can provide a new idea for road background detection of vehicle images captured by highway bayonet.