摘要
SAR影像的独特成像机理导致影像存在相干斑噪声,利用传统的光学影像分类方法难以获得较好的分类结果。鉴于此,提出了一种基于多尺度BP神经网络的SAR影像分类方法,用以对SAR影像进行快速分类。
SAR image has speckle noise due to its unique imaging mechanism. It is difficult to obtain better classification result using the traditional optical image classification methods. On the basis of this, we proposed a SAR image classification method based on multi-scale BP neural network, which could classify SAR images quickly.
出处
《地理空间信息》
2020年第3期72-74,103,I0002,共5页
Geospatial Information
基金
2019年福建省测绘地理信息发展中心科技专项资助项目。