摘要
针对本次项目中每帧视频图像中出现多个矿井巷道灯区域和标示牌区域．提出一种基于连通域标记的矿井巷道灯识别方法。此算法能很好地标记出图像的所有连通区域．并且通过计算各个连通区域的面积特征，很好地把目标识别出来，为下一步机车定位与跟踪打下了基础，此算法满足实时性要求。
In this project, multiple regions and badges miner area appears in each frame of video image. Presents a miner based on connected component labeling for identification. This algorithm can mark the image of all the connected regions, and by calculating the characteristics of each connected domain can identifies a good target, and for its location and tracking further locomotive lays a foundation. This algorithm can satisfy the real-time requirements.
出处
《现代计算机》
2011年第12期24-27,49,共5页
Modern Computer
基金
中央高校基本科研业务费专项(No.JGX101870)
关键词
阈值分割
连通域
连通成分标记
标记图像
Segmentation
Connected Domain
Connected Components Labeled
Mark Image
