基于支持向量机的机场检测算法  

Airport Detection Algorithm Based on Support Vector Machine

作　　者：曲延云[1] 郑南宁[1] 李翠华[2] 

机构地区：[1]西安交通大学人工智能与机器人研究所,西安710049 [2]厦门大学计算机科学系,厦门361005

出　　处：《西安交通大学学报》2006年第6期709-713,共5页Journal of Xi'an Jiaotong University

基　　金：国家自然科学基金资助项目(60205001);国家自然科学基金优秀创新研究群体资助项目(60021302);厦门大学"九八五"二期信息创新平台资助项目

摘　　要：提出了一种新的机场检测算法.该算法通过把机场跑道的几何特征与其所在区域的纹理特征相结合来描述机场特征,其中由灰度的平均值和方差、区域的光滑性、直方图的偏斜度、区域的一致性、图像的随机性、图像的梯度平均和方差等8个特征组成机场的纹理特征向量.先通过直线检测找到机场跑道的候选区域,然后用基于高斯核函数的支持向量机作为分类函数,对候选区域的特征向量进行分类,由此判别机场跑道.实验表明,与传统的仅通过形状判断机场的方法比较,该算法对机场的误检率较低,检测率比刘德红的方法高近10倍,几乎能实时完成一幅图像的检测.A novel airport detection algorithm was proposed, in which the characteristics of the airport was described by combining the texture features and shape features of the runway. Eight texture features including the mean of the region, the deviation of the region, the smoothness of the region, the skewness of a histogram, the uniformity of the region, the randomness of the region, the mean of the gradient image, and the deviation of the gradient image were proposed to form the texture feature vector of the airport. In this algorithm, firstly the straight lines are detected to find candidate regions of the runway, and then the support vector machine based on Gaussian kernel is used as a classifier to classify the candidate regions of the runway and discriminate the runway. The experimental results show that the detection error rate of the proposed algorithm is lower than that of the conventional methods which detect airport only by the shape lea ture of the runway. The detection accuracy of the proposed algorithm is nearly 10 times higher than that of Liu＇s method, and can nearly detect an entire airport image in real-time.

关 键 词：机场检测 支持向量机 直线检测 

分 类 号：TN911.73[电子电信—通信与信息系统]