矿井监控图像中空列车的识别

基于井下空列车与其他物体的几何形状差别,提出空列车监控图像的识别算法.运用圆形结构元素对监控图像进行形态学的先开启、后闭合的操作,用Canny边缘检测算子来检测图像的边缘,并对边缘的图像进行变换分析,突出了空列车图像边缘的直线性质.结果表明：形态学操作有效地减弱了其他对象的边缘,得到比较满意地空列车图像的直线性质的边缘;在不同监控对象的图像的Radon变换域中的特征最大值比Hough变换域中的特征最大值有更好的可分离性.Radon变换作为识别算法中的变换方法,可识别了煤矿井下红外监控系统中的空列车图像.

Recognizing Empty Trains in Coalmine Surveillance Images

Based on the difference of the geometrical shape between the empty trains and other objects. A recognition algorithm of the empty-trains-image was proposed. The monitoring images were opened and closed using a circle structural element in morphological in the algorithm, and their edges were detected with Canny operator. With the transform analysis of the edges images, the linear characteristics of the edges of the empty-trains-images were outstood. The results show that the linear edges of the empty trains were gotten satisfactorily, while that of the other objects were weakened efficiently, and the separability of the diagnostic maximum of the different monitored object in Radon transformation field is better than that in Hough transformation field. The Radon transform can recognize the empty-trains-image in the coal mine infrared monitoring system.