一种应用于交通视频检测的快速边缘提取算法  

A Fast Edge Extraction Algorithm for Traffic Visual Detection

作　　者：范玉强[1] 徐慧[1] 叶雯[1] 

机构地区：[1]南京林业大学信息科学技术学院,江苏南京210037

出　　处：《电子科技》2013年第2期141-143,149,共4页Electronic Science and Technology

摘　　要：车辆边缘信息以及道路边缘信息是交通视频检测中重要的特征信息,而传统的边缘提取算法的运算量大,不能满足交通视频检测实时性的要求。文中突破传统算法中对一幅图像中每个像素点逐一运算的思想,提出了一种快速的边缘提取算法,对一幅图像进行整体运算,大幅降低了边缘提取算法的运算量。通过实验与传统的边缘提取算法比较,证明了该算法应用于交通视频检测效果良好。Edge information of vehicles and roads is of importance in traffic visual detection, but traditional edge extraction algorithms cannot meet the intensive computation requirement in traffic visual detection. To this end, a novel fast edge extraction algorithm is presented, by which each image is treated as a unit in calculation thus reduc- ing the computation amount significantly. Experiments show that this algorithm is more effective in traffic visual de- tection than the traditional ones.

关 键 词：交通视频检测 边缘提取 特征提取 

分 类 号：TP391.41[自动化与计算机技术—计算机应用技术]