结合辐射与分形的高分遥感建筑物阴影检测

Building shadow detection for high resolution remote sensing image combining radiation and fractal

高分辨率遥感影像建筑物阴影检测可为城市摄影测量领域中的建筑物定位、高度估计等提供关键信息。针对传统基于辐射特征的阴影检测方法中所存在的局限,提出了一种结合辐射特征与分形属性的高分辨率遥感影像建筑物阴影检测方法。首先,采用针对高分辨率遥感影像的WJSEG（wavelet J-segmentation）算法进行区域分割;在此基础上,通过综合分析建筑物阴影的辐射及几何分形信息,构建了统一的特征空间;最后,提出了基于用户可信度的样本选择方法来优化主动学习的支持向量机（SVM）分类,从而实现建筑物阴影检测。实验表明,所提出方法能够准确识别建筑物阴影并有效排除虚假目标的干扰,检测精度显著提高。

Accurate detection of the building shadows in high resolution remote sensing image can provide a wealth of key information for the city photogrammetry fields,such as building positioning and height estimation,etc. Aiming at the limitations of traditional shadow detection methods based on radiation features,in this paper a novel building shadow detection method for high resolution remote sensing image combining radiation feature and fractal attribute is proposed. The detection method mainly includes three steps. Firstly,the wavelet J-segmentation（ WJSEG） algorithm aiming at high resolution remote sensing image is adopted to implement region segmentation.Based on above,a unified feature space is constructed through comprehensively analyzing the radiation and geometric fractal information of the building shadow. Finally,a sample selection method based on user reliability is proposed to optimize the SVM classification of active learning; and then the detection of building shadow is realized. The experiment results show that the proposed method can accurately recognize the building shadow and effectively eliminate the interference of the false targets,which significantly improves the detection accuracy.