摘要
提出一种基于特征的机车车载图像配准稳定算法。通过计算视频帧中的特征曲线来估计图像的偏移。算法计算量较低,可以应用快速变化的实时图像序列。对从铁路现场获取的机车车载摄像机视频序列的处理验证本算法的有效性。
This paper researches an effective characteristic- stabilizing algorithm for image sequence from camera set on locomotive. The image dither acquires by characteristic curve lines which are computed from each video frame . This algorithm avoids heavy operation and can be used in real- time image sequence. It is proved that this system is valid to video sequence acquired in railway field.
出处
《铁道机车车辆》
2008年第3期41-43,共3页
Railway Locomotive & Car
关键词
铁路机车
视觉识别
电子稳像
railway locomotive
vision recognition
electronic image stabilization
