【摘要】 利用数学形态学方法,研究与探讨了IKONOS多光谱图像的分割技术,提出一种结合图像边缘特征和纹理特征的混合分割新算法。在高分辨率多光谱遥感图像K-L变换的基础上,采用多尺度多方向形态学梯度算子提取边缘特征,应用数学形态学滤波及局部方差统计特征对图像对象进行标记,最后采用强制最小过程,进行标记控制的分水岭分割。研究结果表明,提出的分割算法优于仅利用边缘特征的分水岭分割算法,同时,该算法能较好地解决分割过程中存在的过分割与欠分割问题,是一种适合高分辨率多光谱遥感图像的分割算法。更多还原

【Abstract】 Image segmentation has been an important research area in image analysis and interpretation.An ideal segmentation strategy of remotely sensed data should consider problems of over-segmentation and under-segmentation simultaneously and find a good tradeoff between them.In this paper,image segmentation for IKONOS multispectral data is investigated by using techniques of mathematical morphology,and a novel hybrid segmentation algorithm is proposed by combining both edge and texture features of images.Based on the K-L transform of multispectral data,edge features are detected by morphological multiscale and multidirection gradient algorithms,and image objects are marked through morphological filtering and local variance features extracting.Finally,the marker controlled watershed algorithm is implemented.The results indicate that the performance of the proposed algorithm is superior to the gradient based watershed segmentation.Moreover,this approach is more suitable for high resolution remotely sensed data to overcome over-segmentation and under-segmentation problems effectively.更多还原