遥感图象上线状目标中心骨架线的提取，是形成图象中层视觉描述的一个重要环节，也是相关领域中具有挑战性的课题之一。为了快速高效地进行遥感图象中线状骨架线的提取，在详细分析遥感图象上线状目标影象特征的基础上，提出了一种半自动提取线状目标中心骨架线的新方法，即变向跟踪法。通过对多种复杂特征的线状地物骨架提取实例表明，提取结果接近于人工判读的结果，且有较好的几何精度和抗干扰性，计算速度也很快。

The problem of liner objects skeleton detection from remotely sensed image is an interesting and challenging one in the correlated field. In order to extract the linear objects skeleton speedily and efficaciously from remotely sensed image, this paper presents a detailed summary of the properties and basic facts about direction changed skeleton tracing method of liner objects based on its various appearance in image. Geometric continuity constraints are of special interest, as they appear to satisfy visual needs of the human observers. Examination is made of the many different liner objects conditions. Experiments detected large numbers of complex properties of linear object skeleton show that the result approximate to manpower distinguish, at the same time this algorithm is better in terms of accuracy, robustness and the computational speed is quite rapid. So it is worth discussing for craft brother.