基于多特征的高分辨率遥感影像道路提取算法研究

发布时间：2018-03-14 01:18

本文选题：高分辨率遥感　切入点：道路提取　出处：《中国矿业大学》2014年硕士论文　论文类型：学位论文

【摘要】：高分辨率遥感是近年来遥感领域发展极为迅速的一项技术，它为对地观测领域提供了丰富、海量的空间数据。道路是高分辨率遥感影像中一项十分重要的特征。从高分辨率遥感影像中提取道路，是高分辨率遥感影像分析、解译和应用的一个前沿课题，，有着重要的理论和现实意义，在现实生活中有非常广泛的应用。虽然已经有大量有关道路提取的算法，但是由于自然场景的复杂性、遥感影像的噪声和算法的局限性，道路提取并没有得到完美地解决，其依然是一个研究的热点和难点。本文从多特征角度出发，研究影响道路提取精度的因素，提出了基于多特征的高分辨率遥感影像道路提取算法，提高了道路提取的精度。研究内容包括道路形状特征、光滑道路中心线提取以及面向对象的道路提取。研究成果为高分辨率遥感影像道路提取提供了一套新的解决方案。本文具体的研究工作主要有：（1）构建了一种道路形状特征指标。本文提出的形状指标可以更加有效地描述道路的几何特征，滤除非线性特征，提高了道路提取精度。（2）提出了一种光滑的道路中心线提取算法。该算法使用张量投票将道路网络分解为独立的道路段，然后利用回归算法从道路段中提取道路中心线。试验结果表明，本文提出的中心线提取算法不会产生类似传统细化算法的毛刺现象，结果比较光滑。（3）提出了一种基于多特征的道路提取算法。该方法融合了形状特征、图像分割和图像分类，试验结果表明该算法具有较高的精度和稳定性。
[Abstract]:High-resolution remote sensing is one of the most rapidly developing technologies in the field of remote sensing in recent years, which has provided a wealth for the field of Earth observation, Massive spatial data. Road is a very important feature in high-resolution remote sensing image. Extracting road from high-resolution remote sensing image is a forward topic in the analysis, interpretation and application of high-resolution remote sensing image. Has important theoretical and practical significance, has very extensive application in the real life. Although there have been a large number of road extraction algorithms, but due to the complexity of natural scenes, remote sensing image noise and the limitations of the algorithm, Road extraction has not been solved perfectly, and it is still a hot and difficult point. From the point of view of multi-feature, this paper studies the factors that affect the road extraction accuracy, and proposes a high-resolution remote sensing image road extraction algorithm based on multi-feature, which improves the accuracy of road extraction. Smooth road centerline extraction and object-oriented road extraction. The research results provide a new solution for high-resolution remote sensing image road extraction. 1) A road shape feature index is constructed, which can more effectively describe the geometric features of the road, filter out the nonlinear features, and improve the road extraction accuracy. (2) A smooth road centerline extraction algorithm is proposed, in which Zhang Liang is used to decompose the road network into independent road segments, and then the regression algorithm is used to extract the road centerline from the road segment. The experimental results show that, The centerline extraction algorithm presented in this paper does not produce burr phenomenon similar to the traditional thinning algorithm, and the result is smooth. (3) A multi-feature based road extraction algorithm is proposed, which combines shape features, image segmentation and image classification. The experimental results show that the algorithm has high accuracy and stability.
【学位授予单位】：中国矿业大学
【学位级别】：硕士
【学位授予年份】：2014
【分类号】：P237

【参考文献】