论文部分内容阅读
针对传统的主动轮廓模型在提取建筑轮廓时不能区分与其具有相似反射率的地物和需要建筑物附近的初始轮廓等问题,从两方面对其进行了改进:1)将原始的单一数据改为利用高分辨率遥感影像与激光雷达(LiDAR)数据融合后的数据进行建筑物的提取;2)在原始模型的能量公式中加入比例系数,控制各波段在能量泛函中的比重.采用Matlab编程实现了所提出的算法,并对徐州市两个地区的快鸟(Quickbird)影像进行了分析.结果表明:改进后的模型可以很好的完成建筑物轮廓的自动提取,并且具有对噪声不敏感、不需要建筑物附近的初始轮廓和隐式改变拓扑结构的优点,达到了较好的效果,证明了改进主动轮廓模型的可行性.所提出的算法为建筑物的轮廓提取提供了有效手段.
The traditional active contour model can not distinguish the objects with similar reflectivity and the initial contour around the building when extracting the outline of the building and so on. It improves the original active contour model from two aspects: 1) The original single data is changed to The building is extracted by using the data fused by high resolution remote sensing image and laser radar (LiDAR) data.2) Proportional coefficient is added to the energy model of the original model to control the proportion of each band in the energy functional. The proposed algorithm is implemented and the Quickbird images in two regions of Xuzhou are analyzed.The results show that the improved model can automatically extract building contours and is insensitive to noise , It does not need the initial contour near the building and the advantage of implicitly changing the topology, which achieves good results and proves the feasibility of improving the active contour model. The proposed algorithm provides an effective means for the contour extraction of buildings.